Wireless communication systems

The present invention relates to a type of wireless communication system in which a base station (BS) communicates with a plurality of fixed or mobile subscriber stations (SS).

In recent years, various standards have been developed for data communication via wideband wireless connection. One such standard was designed with the IEEE 802.16 specification and is commonly known as WiMAX. The specification includes IEEE802.16-2004, which primarily targets systems with fixed subscriber stations, and IEEE802.16e-2005, which specifically specifies mobile subscriber stations. In the description below, the term Subscriber Station (SS) applies to both fixed and mobile stations (SS / MS).

The entire contents of the IEEE standard 802.16-2004 "Air Interface For Fixed Broadband Wireless Access Systems" are incorporated herein by reference. IEEE802.16 envisions a single-hop system in which a subscriber station communicates directly with the base station within at least one "cell" defined by the base station. By deploying base stations in appropriate locations within a given geographic area and / or by providing multiple antennas on the same base station, adjacent groups of cells are created to form a wide area network. obtain. As used herein, the terms "network" and "system" are used interchangeably.

In the above types of systems, data is transferred by packet switching between the subscriber station and the base station while the connection (management connection or transport connection) is maintained between the subscriber station and the base station. Be communicated. The direction in which packets are transmitted from the subscriber station to the base station is called the uplink (UL). The direction from the base station to the subscriber station is called the downlink (DL). Packets have a defined format according to the hierarchical protocol applied to the system and the radio equipment of the components of the system. The layers of the protocol for packets are themselves the so-called physical layer (PHY) and media access layer (MAC). In the IEEE802.16-2004 specification, these protocol layers form the protocol stack shown in Figure 1.

The media access layer shown in FIG. 1 is the protocol layer most relevant to the present invention described herein. The media access layer is involved in the processing of various functions including network access, bandwidth allocation and connection maintenance. This includes access control of BS and SS to the network based on "frames" that are divided into a large number of slots in a predetermined unit time and time domain in the system. Data is exchanged between MAC peer entities, in other words, between subscriber stations and base stations, in units of protocol data units (PDUs). PDUs are propagated across the PHY layer using multiple slots. Therefore, a "slot" is a unit of time used to allocate bandwidth. The MAC is divided into sub-layers with security sub-layers that allow authentication, key exchange and PDU encryption (see Figure 1). These functions and sub-layers can be grouped into a "MAC upper layer" or UMAC with generally higher levels of functionality and a MAC lower layer or LMAC with lower levels of speed-focused functionality.

UMAC is typically performed by running software on a general purpose processor, allowing reconfiguration when updates and system changes are requested. The functions of UMAC include MAC management, service-dependent convergence layer and MAC common sub-layer (MAC CPS, see Fig. 1) as shown in Fig. 1. The LMAC may also be provided by software executed by the processor, but this example requires low-level code (possibly implemented on the processor) and / or a real-time operating system. The LMAC acts as a bridge between the UMAC and the PHY, data encryption / decryption (a security sublayer feature shown in Figure 1), error correction code (CRC) generation, PDU classification, and further described below. Offload some of the UMAC tasks by performing block processing.

Given the transmission of data, data flow generally flows from top to bottom of the protocol stack. Thus, for example, data packets (or so-called service data units, SDUs, described in more detail below) are from the top of the protocol stack (application layer, not shown in Figure 1) to the CS in Figure 1. Sent via SAP. In UMAC, SDUs are queued to LMAC and converted to MAC PDUs (in a process called "packing") to form subframes. So-called "fragmentation" is also required because the data sizes of SDUs and PDUs do not have to match. With fragmentation, a single SDU is split among multiple MAC PDUs. At the physical layer, the assembled subframe is prepared for transmission by assigning the components of the subframe to one of many "bursts" of electromagnetic waves. Figure 2 schematically shows the relationship between SDU, PDU and burst.

In burst, forward error correction (FEC) is used to help the receiver correct errors introduced by the transmit process. Each burst can have multiple FEC blocks as shown in FIG. The MAC PDU is contained within the FEC block, and the MAC PDU may span multiple boundaries of the FEC block.

Various physical layer implementations are possible in an IEEE 802.16 network, depending on the available frequency range and application. For example, the Time Division Duplex (TDD) scheme and the Frequency Division Duplex (FDD) scheme described below. The PHY layer also defines transmission techniques such as OFDM (Orthogonal Frequency Division Multiplexing) or OFDMA (Orthogonal Frequency Division Multiplexing Access). These techniques are briefly described below.

In OFDM, a single data stream is modulated into N parallel subcarriers. The signal of each subcarrier has its own frequency range. It divides the entire bandwidth (ie, the amount of data to be transmitted at a given time interval) across multiple subcarriers, thereby increasing the duration of each data symbol. Since each subcarrier has a low information rate, a multicarrier system is advantageous because it introduces less distortion into the channel than a single carrier system. This is made possible by ensuring that the transmit rate, and thus the bandwidth of each subcarrier, is less than the coherence bandwidth of the channel. As a result, the channel distortion suffered by a single carrier is frequency independent and can therefore be corrected by simple phase and amplitude correction factors. Therefore, correcting channel distortion within a multicarrier receiver can significantly reduce the complexity of opposites within a single carrier receiver when the bandwidth of the system exceeds the coherence bandwidth of the channel.

OFDM systems use multiple subcarrier frequencies. Since multiple subcarrier frequencies are orthogonal in a mathematical sense, the fact that the spectra of the subcarriers are independent of each other causes the subcarriers to overlap without interference. The orthogonality of an OFDM system does not need to protect the band frequency, thereby improving the spectral efficiency of the system. OFDM has been proposed and applied to many wireless systems. The OFDM system uses an inverse discrete or fast Fourier transform algorithm (IDFT / IFFT) to map the signal blocks of N modulated parallel data sources to N orthogonal parallel subcarriers and time in the transmitter. It forms a signal known as the "OFDM symbol" of the region. Therefore, the "OFDM symbol" is a composite signal of all N subcarrier signals. At the receiver, the time domain signal received is converted back into the frequency domain by applying a Discrete Fourier Transform (DFT) or Fast Fourier Transform (FFT) algorithm.

OFDMA (Orthogonal Frequency Division Multiplexing) is a multiple access version of OFDM. OFDMA works by assigning a subset of subcarriers to individual subcarriers. This allows simultaneous transmission from multiple users, resulting in even better spectral efficiency. However, there is still a problem of realizing two-way communication, that is, enabling communication without interference in the upstream and downstream directions.

There are two well-known and different techniques for achieving two-way communication between two nodes. These techniques double the two (forward or downlink, and reverse or uplink) communication links to overcome the physical constraint that the device cannot transmit and receive simultaneously on the same resource medium. The first method, Frequency Division Duplex (FDD), operates two links simultaneously, but uses two separate bands of transmission medium, one for DL communication and one for UL communication. Operates in different frequency bands by dividing into. The second technique, Time Division Duplex (TDD), operates two links in the same frequency band, but divides simultaneous access to the medium, and at any given time only DL or UL use the medium. The two methods (TDD and FDD) have their respective advantages and are both commonly used in single-hop wired and wireless communication systems. The IEEE 802.16 standard includes both FDD and TDD systems, but the specification of the present application mainly describes the TDD system below.

Figures 3 and 4 show the TDD frame structure used in the OFDMA physical layer mode of the IEEE 802.16 standard (WiMAX).

In FIG. 3, the frame is a given time length and a given frequency band, that is, the time axis shown by "OFDM symbol number" in FIG. 3, and "number of subchannels" (each subchannel is a set of the above-mentioned subcarriers). Is considered to occupy the frequency axis indicated by). Each frame is divided into DL and UL lower frames, which are separate transmission intervals. They are separated by transmit / receive and receive / transmit transition protection intervals (TTG and RTG, respectively). Each DL subframe begins with a preamble, followed by a frame control header (FCH), DL-MAP, and UL-MAP. The FCH contains the DL frame prefix (DLFP) and specifies the burst characteristics and DL-MAP length. The DLFP is the data structure sent at the beginning of each frame and contains information about the current frame that is mapped to the FCH. Simultaneous DL allocations are broadcast, multicast and unicast, which include allocations for another BS rather than servicing the BS. Simultaneous UL may be data allocation and ranging or bandwidth requirements.

Figure 4 illustrates the OFDMA TDD frame structure from different perspectives with two parts within the UL subframe: the PHY header and the MAC PDU. The MAC PDU also has a MAC header, any payload and any error correction code (Cyclic Redundancy Code or CRC). The PHY layer header contains training sequences, frequency band allocation information and other information regarding physical layer parameters. Within a MAC PDU, the MAC header usually provides essential parameters for medium access such as PDU type, MAC address and MAC signal type. The CRC in the MAC PDU is optional and can be used to inspect the received MAC PDU. The payload in the MAC PDU is used to contain the data that the SS wants to send to the BS, but is optional. For example, a control message or ACK message such as a bandwidth request does not have any payload. The payload can be a sub-MAC header that can provide higher layer data or additional MAC information.

In addition, 802.16e OFDMA provides subchanneling as a better means of managing network performance, solving dependent convergence layers and capacity requirements. The OFDMA physical layer divides the available OFDM symbols and component subcarriers (see Figure 3) into separate logical and physical subchannels, with multiple bursts coexisting or as shown in Figure 3. Allows overlap at each time interval. On the downlink, a single burst may be shared by multiple users (subscriber stations), but on the uplink, each burst generally corresponds to a single user. OFDMA subchanneling techniques include frequency division and frequency selective transmission schemes.

Frequency division transmission schemes can be grouped into Full Usage of Subchannels (FUSC) and Partial Usage of Subchannels (PUSC) modes. These modes correspond to frequency division transmission. In frequency division multiplexing, the subcarriers assigned to each logical subchannel are pseudo-randomly distributed over the set of available subcarriers. In FUSC, subcarriers are distributed over the entire frequency range, while in PUSC, several distributed clusters of subcarriers are used to form subchannels. These schemes are more suitable for dealing with changing channel conditions and provide frequency diversity that benefits network coverage and capabilities.

Frequency selective subchanneling is supported in Adaptive Modulation and Coding (AMC) mode. Bandwidth AMC allows subchannel configurations through physically adjacent subcarrier arrangements, or contiguous groups of subcarriers. The system scheduler can utilize closed-loop channel feedback technology to determine the optimal subchannel to be assigned to each subcarrier based on unique channel conditions. FIG. 6 shows an OFDMA TDD mode frame configuration with FUSC, PUSC and AMC zones. In general, FUSC and PUSC are suitable for connections between base stations and mobile stations. On the other hand, AMC is suitable for connecting to fixed subscriber stations.

Subchanneling is important for frequency reuse schemes that allocate frequencies between adjacent cells. Perhaps the most common reuse scheme is referred to as "reuse 3" (reuse factor 3). In this scheme, to reduce interference, hexagonal cells are considered to have each pair of adjacent cells assigned a different set of frequency channels. Three sets of channels are sufficient to achieve this. PUSC or FUSC is a suitable transmission scheme in this case. This is because the random allocation of subcarriers to subchannels further minimizes the possibility of interference between signals in different cells. On the other hand, "reuse 1" simply means using the same frequency set within each cell (reuse factor 1). This increases interference (decreases CINR), but allows the entire available frequency range (all subchannels in Figure 3) to be used by each connection, and is simple enough to be in a real system. Can be implemented. It is possible to use different reuse schemes for different subcarriers in a single cell at the same time. In particular, Reuse 3 is suitable for users near the edge of the cell, while Reuse 1 can be safely used for subcarriers near the center of the cell where interference from other cells is unlikely to occur. This results in an "efficient reuse factor" for any system between 1 and 3. Normally, each cell is provided with a base station located in the center of each cell, but gradually, antennas in multiple directions are attached to a single base station, and the same base station serves multiple cells around the base station. It can be provided.

The DL subframe contains a broadcast control field with DL-MAP and UL-MAP. As a result, the BS informs the receiving device of the frame structure. A MAP is a map of bandwidth allocation within a frame, including an information element (IE), each with a connection ID. IE on the map informs the subscriber station of the burst that the subscriber station is assigned to receive the information. Therefore, in a TDD system network, bandwidth allocation means allocation of resources (slots) within a frame. DL-MAP and UL-MAP are examples of BS broadcasting management messages (that is, sent to all subscribers). Other management messages include uplink channel descriptor UCD and downlink channel descriptor DCD (both shown in Figure 4), dynamic service requests and responses (DS-REQ and -RSP).

The concept of quality of service (QoS) is used in wireless communication systems to enable the provision of a wide range of services. Depending on the type of service provided (see below), the packet should be sent with some accuracy and / or within a certain time delay. Alternatively, the packet is corrupted and may need to be retransmitted. During communication with a subscriber station, the base station keeps in mind that the base station typically communicates with multiple subscriber stations at the same time, depending on the type of service requested by the subscriber station and the available bandwidth. And assign a QoS level. The QoS parameters take into account transmission priority (time delay or latency), transmission accuracy (error rate), and throughput (data rate).

The connection between the base station and the subscriber station (more specifically, between the MACs of those devices, the so-called peer entities) is assigned a connection identifier (CID) so that the base station manages the active connections of the base station. Keep track of CID. In order to support address assignment and QoS control, some wireless communication systems add a connection identifier (CID) to the MAC header. For example, in WiMAX, service flows between SS / MS and BS can be created and activated during network entry procedures or by dynamic service flow procedures. As mentioned above, a service flow ID (SFID) is assigned to each existing service flow. Each service flow is also associated with a particular QoS request. The service flow has at least one SFID and associated directions. The transport connection's connection identifier (CID) exists only if the service flow is authorized or active. The relationship between SFID and transport CID is unique. This means that the SFID should not be associated with more than one transport ID and that the transport CID should not be associated with more than one SFID.

BS uses a scheduler (scheduling algorithm) to manage the allocation of bandwidth (slots) to all currently active connections and balance the demands of various subscribers. That is, each SS only needs to negotiate once for a network entry. The bandwidth allocated by the BS then increases or decreases at the request of the SS or due to other needs of the network, but remains allocated to the SS and thus keeps the connection active.

Whenever possible, the scheduler ensures that all connections that need to be serviced within the current frame (especially DL subframes consisting of base stations) receive some resources (bandwidth). There must be. Apart from this QoS requirement, other factors to consider include the distance from the base station to each subscriber (route loss) and, if necessary, the mobility of the subscriber. If the subscriber is far or far from the base station, the transmission rate achievable on the downlink is more efficient (from a system performance standpoint) to preferentially treat subscribers closer to the base station. Is reduced to. On the other hand, no subscriber can tolerate the loss of data.

A technique aimed at balancing these factors is called the "proportional fair (PF)" algorithm. The algorithm aims to achieve a balance between system performance and fairness to subscribers by maximizing the logarithm of the long-term average data rate provided to all subscribers.

As mentioned above, each connection has a QoS associated with the class of service. QoS is first assigned during the network entry procedure (connection establishment stage) when a subscriber station joins the network, and then a subscription that makes a request to the base station while the connection is maintained. It may be changed by the network. This will probably repeat, with additional bandwidth allocation to the connection, depending on the resources available in the network.

The relationship between QoS and CID / SFID is illustrated in Figure 5. To simplify the understanding of Figure 5, it should be noted that the "service flow" is the transmission of data in a given direction (uplink or downlink) over a connection with a particular QoS. To show. The QoS of a connection is defined by a service flow identifier (SFID) that has a one-to-one relationship with the connection identifier. Strictly speaking, bandwidth is allocated to the service flow (or connection), but it is convenient to consider the bandwidth allocated to the SS included in the connection by BS. Each service flow can be classified as one of a set of service classes or QoS classes. Although basically the same for both DL and UL, these service classes are defined to be slightly different from the perspective of the DL and UL schedulers. There are differences between DL and UL QoS delivery mechanisms. This difference is due to the fact that the BS cannot directly see the buffer state in the MS and may not know the packet error rate in the MS.

<Downline (DL)> The following service class types are defined by IEEE 802.16. -UGS: Unsolicited grant service -RT-VR: Real-time variable rate service -ERT-VR: Extended Real-time variable rate service -NRT-VR: Non Real-time variable rate service -BE: Best Effort service.

Table 1 provides a brief description of the objectives and lists the parameters associated with each service class. These parameters are described in Table 2. [Table 1] Overview of DL service types<tables num="1"><img file="JP4998564B2_D0001.tif" /></tables>

[Table 2] Parameter description<tables num="2"><img file="JP4998564B2_D0002.tif" /></tables>

The role of the downlink packet scheduler is to ensure that the requirements set for each active service flow are met based on the set parameter settings.

&lt;Uplink (UL)&gt; The following service class types and scheduling services are defined by the IEEE802.16 standard and are supported by the WiMAX Forum Mobile System Profile. -UGS: Unsolicited grant service -rtPS: Real-time Polling Service -ertPS: Extended Real-time Polling Service -nrtPS: Non-Real-time Polling Service -BE: Best Effort service

Table 3 provides a brief description of the objectives and lists the parameters associated with each service class (parameter descriptions are shown in Table 2). [Table 3] UL Service Type Overview<tables num="3"><img file="JP4998564B2_D0003.tif" /></tables>

The role of the uplink packet scheduler is to ensure that the requirements set for each active service flow are met by appropriately allocating resources to the MS CID based on the configured parameter settings. .. Especially in the case of UL, the scheduler has the opportunity to poll for a service flow with a polling interval (eg rtPS or nrtPS) or a UGS connection with a PM bit set or an unsolicited permission request based on the flow parameters. Sufficient BW) must also be provided to send the BW request.

UL also requires the scheduler to review the information received from the MS regarding the current requirements of the MS. These requirements include: -PBR, PM, SI, FL and FLI bits in the authorization management subheader (see below) -BR (additional or aggregate) in MAC signaling type I header -CDMA BW request

Further details of these information items are shown in the table below.<tables num="4"><img file="JP4998564B2_D0004.tif" /></tables>

&lt;Corresponding classes and parameters&gt; Table 4 shows which parameters are required, optional, or inapplicable for each service class. [Table 4] Service class parameters corresponding to each type<tables num="5"><img file="JP4998564B2_D0005.tif" /></tables>

<p> As is clear from the above description, the tasks of the UL and DL packet schedulers in the base station are highly relevant. However, in WiMAX systems, base station functionality must be offered at a low price in order to be commercially successful. Therefore, it is necessary to provide the above-mentioned scheduling function by a simple method so as to minimize the required processing capacity and its price.</p><p> More specifically, there is a need for subsystems that provide effective and cost-effective LMAC schedulers and PDU formation.</p>

<p> According to the first aspect of the present invention, a module used in a base station of a wireless communication system is provided. In the wireless communication system, at least a packet of data is transmitted from the base station to a plurality of subscriber stations, the base station defines a connection with the subscriber station, and each connection is of a plurality of service classes. Having one, each class of service has associated QoS, the module allocates system resources available to each connection within a given unit time, and the module: said QoS requirements for the connection. A first-stage scheduling means that determines which connections need to be scheduled within the predetermined unit time to achieve QoS for the connection; resources are allocated to the connections that need to be scheduled. A second-stage scheduling means that later uses resources that are still available to determine other connections that can be scheduled within the predetermined unit time and prioritize the other connections based on non-QoS requirements; and said. It has a first-stage scheduling means and a resource allocating means that allocates resources to each connection according to the decisions made by the second-stage scheduling means;</p><p> In the module described above, requirements other than the QoS are defined by the operator of the wireless communication system. In order to change the method of scheduling the other connections, the module may have a means of receiving information about requirements other than QoS. The purpose of the second stage scheduling means may be to maximize the spectral efficiency of the system. Alternatively, a proportional fair algorithm may be used to have fairness to the subscriber and balance the spectral efficiency.</p><p> According to the second aspect of the present invention, a module used in a base station of a wireless communication system is provided. In the wireless communication system, a frame of data is transmitted from a plurality of subscriber stations via the base station, the frame is divided into a plurality of zones in time, and each zone has a plurality of bursts. The base station receives signals from subscriber stations in a plurality of adjacent regions, establishes a connection with the subscriber station, allocates resources in the frame to provide services to the connection, and provides the module. A zone allocation means that receives a signal level for each connection and performs zone allocation for the connection based on a comparison of the signal level with the threshold level; and said threshold level used by the zone allocation means for zone allocation. Has a partial frequency reuse control unit;</p><p> The module is used, for example, in a base station having a three sector antenna to serve adjacent hexagonal cells such as the region. In the above configuration, preferably, the zone has 1 reuse zone and 3 reuse zones, where 1 and 3 indicate the frequency reuse coefficient between the cells, and reuse 1 indicates the frequency reuse between the cells. Indicates no utilization, reuse 3 indicates that the frequency is divided into three sets of subchannels within the frequency band occupied by the zone, each set being assigned to a different cell and said threshold level. Is the optimal reuse 3 / reuse 1 transition signal level.</p><p> Another aspect of the present invention is a subsystem having one or both of the above modules, a base station having the module or subsystem, a scheduling device, a scheduling method, a wireless communication method, a radio, as defined in a dependent claim. Provide software that implements the functions of a communication system and the modules defined above.</p>

<figref num="1">The protocol layer according to IEEE802.16 is shown.</figref><figref num="2">The relationship between SDU, PDU and burst in the IEEE802.16 wireless communication network is shown.</figref><figref num="3">The TDD frame format used in the IEEE802.16 wireless communication system is explained, and the frame structure from the viewpoint of burst is shown.</figref><figref num="4">The TDD system frame format used in the IEEE802.16 wireless communication system is explained, and the frame structure from the viewpoint of PDU is shown.</figref><figref num="5">The relationship between CID, SFID and QoS in the IEEE 802.16 network is shown.</figref><figref num="6">The modified frame format in the case of frequency reuse is outlined.</figref><figref num="7">The overall structure of a base station subsystem that implements the present invention is shown.</figref><figref num="8">It is a flowchart of the whole operation of the subsystem of FIG.</figref><figref num="9">The first aspect of the operation of the subsystem of FIG. 7 with respect to packet scheduling is shown.</figref><figref num="10">The frame format used in the so-called reuse 3 / reuse 1 base station configuration is shown.</figref><figref num="11">The relationship between the spectral efficiency of the reuse 3 / reuse 1 configuration and the effective reuse factor is shown.</figref>

See the attached drawing as an example.

&lt;Mode for implementing the present invention&gt; Examples of the present invention are described with reference to FIG. 7-11, using the IEEE 802.16 network as an example. First, a high-level description of the LMAC scheduler and PDU forming unit that enables WiMAX base stations to guarantee QoS delivery and maximizes spectral efficiency will be described. In short, this is achieved through a combination of packet scheduling based on the connection of queued packets based on service flow type, related parameters and channel state considerations. Next, some algorithms included in each component related to the scheduler and the PDU forming unit will be outlined.

Figure 7 shows the high-level architecture of the WiMAX base station's LMAC scheduler and MAC PDU formation for downlinks. It should be noted that the UL scheduler does not require the elements that form the data plane, as described below. However, in reality, the same processor will be used to implement both the DL and UL schedulers, and only the settings will be changed.

&lt;Downline (DL) Scheduler&gt; The components of the downlink scheduler and PDU forming unit are divided into data or control plane parts. The data plane has the following components: -Classification section (of incoming packets) -SDU Queue Management Department -SDU dequeue department -PDU forming part (inside the subframe forming part) -PHY interface The overall behavior of the components of the data plane is to queue the incoming (input) SDU and convert it to an outgoing (output) MAC PDU for transmission to the PHY layer. To achieve this, the function of the PDU forming unit requires information from the function of the burst map unit that informs how the PDU forming unit operates.

The control plane has the following components. -Service Flow Management Department -Subframe decision section -SDU Queue Management Department -FFR management department -Station management department -Connection scheduler -Burst map part (inside the subframe formation part) The overall operation of the components of the control plane is to provide the PDU-forming unit with the inputs it needs to perform its function to the PDU-forming unit in the data plane and, as a result, to send it to the PHY layer. To build a MAP message with IE. This is ultimately provided by the burst map section. The burst map section relies on information from other components of the control plane, as shown.

&lt;Uplink scheduler&gt; Although not shown, the functional components in the uplink scheduler have some of the components in the downlink scheduler. In the case of the uplink, the BS has no data plane, instead the equivalent of the DL component exists in the MS. Therefore, the UL architecture has only a burst map section of a service flow management section, a station management section, a connection scheduler, and a subframe forming section.

Figure 8 shows a flow chart-based representation of the connection scheduling and PDU formation algorithms at the system level. The operation of each block in the flowchart will be described in detail below.

&lt;Description of high-level components&gt; (i) Bureau management department At each approved subscriber station, the station management department maintains a data structure with station status information. The state information itself has a data structure that has state information for active management and transport connections. At a minimum, the information contained in the structure is needed by the connection scheduler and the frame forming unit.

The station management department also has many functions that affect the data structure. The function includes: -Zone assignment: Assign stations the appropriate zones within the subframe. -Burst profile assignment: Assign a burst profile to a station. -MS / CID status update: Update the station / connection status data structure.

The MS / CID state update function is started when new station information is received (for example, when a service flow change, CINR report is received, a connection is scheduled, or a new SDU is received). It may be executed periodically and fetch and update information such as SDU queue length or latency information.

It should be noted that as an alternative, the SDU information is from the Burst Map Department and the SDU Queue Management Department, theoretically updated from the SDU Queue Management Department when a new SDU arrives or the SDU is scheduled. It can be tracked based only on the information that needs to be done. Alternatively, the station management unit may increase the waiting time for each frame. This information is updated based on the event and requires explicit update information from the SDU queue management unit or is periodically updated based on periodic requests to the SDU queue management unit.

Based on the new received information, the function initiates the zone assignment or burst profile assignment function and updates the current assignment.

The zone allocation function and burst profile allocation function handle only SISO users. With the goal of designing a framework for DL MIMO users (which needs to support Wave2), there are two techniques to enable MIMO assignment: -Set separate SISO and MIMO zones using STC_Zone / DL_Zone_Switch_IE: The subframe determiner must set separate zones. -Set a MIMO burst in a SISO zone using MIMO_DL_Basic_IE or MIMO_DL_Enhanced_IE: The subframe determiner should set the burst for MIMO allocation and the burst map should allocate space from the burst to the scheduled connection for each MIMO. There is.

(ii) Connection scheduler The connection scheduler determines which connections should be scheduled for each frame.

This decision, which results in a list of connection descriptors (CDs), is based on the number of input parameters available from the station management department. The overall operation is aimed at ensuring QoS as described by the active parameter set and maximizing spectral efficiency.

The first function is to determine which connections should be scheduled with high priority to ensure that the guarantee of maximum QoS is met. These connections are determined by extracting connections that do not meet the minimum QoS requirements if not scheduled within the current frame. This first part begins by analyzing the connection based on the class of service. Analyze RT, ERT, and finally NRT connections, starting with UGS connections. Once scheduling to meet the minimum QoS is complete, the second step is to create a ranked list of the remaining connections and consider allocating resources within the frame. This ranked list is then attached to the extracted list of connections as part of the first step.

The connection scheduler then passes a list of connection descriptors, which are the ranked data objects, to the burst map section. These data objects have the information that the burst map section needs to determine the size of the grant that will then be delivered to the PDU forming section.

Figure 8 details the method used to reach the list of ranked connection descriptors. The high-level flowchart shown in Figure 8 shows the two steps performed by this component and the resulting connection descriptor structure.

(iii) SDU dequeue section The SDU dequeue unit supplies the SDU (or fragment) to the PDU forming unit on demand, fetches the SDU (or fragment) from the appropriate SDU queue for the requested connection, and operates the PUD forming unit. Holds the SDU fragment created as a result of. It should be noted that the SDU dequeue unit, SDU queue management unit and classification unit are designed so that incoming packets can be processed by any connection with changing priority.

(iv) Classification Department and SDU Queue Management Department The classification unit takes in the incoming SDU and places the SDU in an appropriate buffer. The SDU queue management department monitors the queue length and SDU waiting time, and notifies the station management department of this information upon request. The SDU queue management department also notifies the station management department of the arrival of new data in the previously empty queue. This latter function allows the station management to update the MS / CID state when data becomes stagnant in the SDU buffer.

(v) Subframe forming part The subframe forming part has two components, that is, a burst map part and a PDU forming part, and one data structure (subframe resource map) described below.

The subframe former sequentially resolves the entire list of ranked connection descriptors supplied by the connection scheduler until one of the following termination conditions is met: -The deadline for building subframes has been reached. -You have reached the end of the CD list. -There are no resources left in the burst. -There are no resources left in the space allocated for MAP messages.

The burst map section determines the target burst in the allocated zone to which the PDU of the scheduled connection is mapped. The burst map section also calculates the allocation size within the burst based on the number of input parameters, as described in detail below. Once mapped, the burst map section updates the subframe resource map to indicate the resources used within that subframe. The burst map part also updates MAP IE in the case of DL and creates IE in the case of UL. These IEs are stored in the downlink subframe resource map.

The burst map unit notifies the station management unit of resource allocation. This allows the station management department to update the appropriate elements in the MS / CID state data structure. The burst map section also informs the PUD form section of the size of the PUD to be formed on this connection and other supporting information.

The PUD forming unit forms a PUD to the size specified by the burst map unit. When determining the size of the payload available for the SDU, the PUD forming unit sets the PDU size for the MAC header, subheader and CRC. The PUD forming unit notifies the SDU dequeuing unit of the maximum size of data that the PUD forming unit can accept in the payload. The SDU dequeue section then takes into account the overhead required to include the required fragments or packs the subheaders (ie 8 or 16 bits depending on the extension type and the use of the ARQ) SDU (possibly 8 or 16 bits). Returns (with fragments at the beginning or end). The PUD former then constructs the PDU and adds the CRC where appropriate.

(v-1) Subframe resource map This is a data structure with a logical map of subframes. This map contains areas for control (eg, MAP, ranging, CQICH, ACK, etc.) and areas for transport traffic (eg, data) defined by the subframe forming section described below. Has a burst).

In the case of DL subframes, the subframe resource map has a hierarchical map of the zone and bursts within the zone defined for the subframe by the subframe former. Also, the composition of MAP messages, such as assignments, should be tracked by the burst map section. In the case of DL, most of the contents of MAP IE are predetermined. However, when the burst map section makes in-burst allocations, the burst map section updates the CID list in the MAP IE contained in the subframe resource map.

In the case of UL subframes, the subframe resource map has a hierarchical map of the zones defined for the subframe. You can also access part of the DL subframe resource map with UL MAP messages. When the burst map section allocates bursts, the burst map section inserts the MAP IE into the UL MAP contained in the DL subframe resource map and updates the use of the appropriate data zone.

(vi) Service Flow Management Department The service flow management department manages the authorization and activation of the service flow. The management is performed using a permit control algorithm. The main function of the service flow management department regarding the scheduler is to notify the station management department of the addition, change or deletion of the active service flow.

The service flow management department is responsible for flow permission control. The authorization control algorithm may request feedback on the availability of surplus radio resources from the station management or connection scheduler. The service flow management unit determines and manages the settings of QoS parameters set for each service flow. Next, when the service flow is started, the service flow management department notifies the station management department of these values.

(vii) Subframe determination section The subframe determination unit defines the layout of the subframe and stores it in the subframe resource map of the subframe construction unit. The subframe determination unit prepares resources for control messages including the following. -FCH / MAP in DL subframe -CQICH, ACKCH, ranging in UL subframes The subframe determination unit also prepares other areas to be used by the PHY layer. For example, a sounding zone, PAPR reduction or gap zone (DIUC or UIUC13). In conjunction with the network management entity responsible for coordinating the use of resources throughout the system, the subframe determiner also determines the number and location of zones within DL and UL subframes. Finally, within the DL zone, the subframe determiner contains a MAP message containing a DL MAP IE with an empty CID list where the burst map section starts allocating resources before the burst map section starts allocating. Before configuring, define the burst layout.

The subframe determinant either monitors the subframe resource map frame by frame or communicates with the burst map to determine the current use of bursts, zones and subframes and generate local changes. Report on possible system-wide changes to be achieved for network management entities. Therefore, in the case of DL, this layout can be dynamically adjusted based on the monitoring of the current use of bursts in subframes and the list of burst profiles defined in the DCD message.

(viii) FFR Management Department FIG. 11 shows the configuration of DL subframes corresponding to frequency division multiplexing (FFR). The zone transition point between the "Reuse 3" and "Reuse 1" zones is coordinated throughout the network as shown in Figure 11, which causes the subframe decision unit to work with the network management entity for Reuse 3 and Reuse 1. It is expected to set a zone transition point. The role of the FFR management department is to optimally select the thresholds used by the zone allocation function of the station management department and determine whether the MS is assigned to the "R3" or "R1" zone. This allows the system to operate in the optimum operating area as shown in FIG. Further details regarding the zone allocation algorithm and the threshold calculation algorithm will be described below.

&lt;Explanation of LMAC scheduling components of the PDU generator&gt; The main components of the WiMAX BS scheduler and PDU generator will be described in detail below.

(a) Bureau management department The station management department has three functions and a large number of data structures. These will be described separately in the following categories.

(a1) MS state data structure This structure maintains the current state of parameters that are specific to MS but apply to all MS connections (or service flows). The parameters are shown below.<tables num="6"><img file="JP4998564B2_D0006.tif" /></tables>

The MS state data structure also has a CID state object for each active connection.

The BS and MS PHY layers may support only physical CINR reports (Wave 1) or additionally support effective CINR reports (Wave 2). The MS state data structure can be extended to have either physical CINR measurements or effective CINR measurements, as well as physical CINR.

(a2) CID state data structure This data structure is used to maintain the current state of parameters specific to MS connections (or service flows). The parameters are shown below.<tables num="7"><img file="JP4998564B2_D0007.tif" /></tables>

The Maximum Allowed SDU Latency is calculated based on the Maximum latency set in the QoS parameter set. Maximum latency is the CS peer-to-peer latency. Therefore, the Maximum Allowed SDU Latency needs to take into account the special delay between SDU scheduling and arrival at the CS peer. The station management department performs this calculation of the Maximum Allowed SDU Latency.

(a3) MS / CID status update This function updates the data structure of the MS status (MS Status) and the related CID status (CID Status) based on the event. The events that cause the update are shown below.<tables num="8"><img file="JP4998564B2_D0008.tif" /></tables>

(a4) CQI Report Event When the CQI report for the pilot CINR is received, the average pilot CINR (Average Pilot CINR) is updated. The formula used for the update is shown below.<maths num="1"><img file="JP4998564B2_D0009.tif" /></maths>Here, CINRav [k] is the average pilot CINR in frame k, CINR [k] is the pilot CINR in frame k, and β is a forgetting factor that should be set appropriately.

Simulations are used to reach the optimum setting of β with the aim of ensuring that the average CINR used by the proportional fair algorithm is appropriate. If the average window is too long, the proportional fair algorithm becomes like a round robin and does not provide any spectral efficiency benefit. On the other hand, if the average window is too short, the proportional fair algorithm will look like a maximum C / I and will not provide any fairness.

(a5) Successful authorization / polling assignment event If the grant is successful, the Last Grant Time is set equal to the current time. The SDU Buffer Size or Requested BW is decremented by the number of SDU bytes allocated. If the service has an unsolicited grant interval, the Next Grant Time is set based on the sum of the current time and the unsolicited grant interval. Finally, if the service has a set Next Polling Time, the next polling time is set based on the sum of the current time and the Unsolicited Polling Interval. If the polling assignment is successful, the Next Poll Time is set equal to the sum of the current time and the unsolicited polling interval.

(a6) Zone placement (FFR) Whenever a preamble CINR report is received, the zone assignment function is invoked by the MS / CID update function. The operation of the zone allocation function depends on whether the MS is located in the reuse 1 or reuse 3 zone.

Here, CINR1 is used to represent the CINR measurements made to the preamble for reuse configuration 1. CINR3 also represents the CINR measurements made for the preamble for reuse configuration 3.

(a7) Operation of reuse 3 If the MS is currently assigned to a reuse 3 zone, the zone allocation feature checks to see if the CINR1 report is above the R1 / R3 transition threshold. Above the threshold, the zone assignment function assigns the CID to the R1 zone. If not above the threshold, the connection remains in the R3 zone. The zone allocation function then schedules the PHY layer and performs the next periodic check of CINR1.

(a8) Reuse 1 operation (reception of CINR1 report) For reuse 1 zones, the zone allocation feature checks to see if the CINR1 report is below the R1 / R3 transition threshold. Below the threshold, the zone allocation function requires a CINR3 measurement. If not below the threshold, the connection remains in the R1 zone. The zone allocation function then schedules the PHY layer and performs the next periodic check of CINR1.

(a9) Reuse 1 operation (reception of CINR3 report) If the CINR3 report is higher than the CINR1 report, the zone assignment feature assigns the CID to the R3 zone. The zone allocation function then schedules the PHY layer and performs the next periodic check of CINR1. One possibility of getting a report is to set up a periodic CQI report via CQICH for the CINR1 report. The REP-REQ / RSP can then be used to obtain CINR3 reports on demand.

(a10) Burst profile assignment Whenever a pilot CQI report is received, the burst profile assignment feature is initiated by the MS / MS / CID update feature to ensure proper burst profile assignment for MS connections. Since MS only supports physical CINR measurements for pilot subcarriers in Wave1, the burst profile assignment algorithm uses this metric. The burst profile assignment function determines the optimal burst profile format from the list of active burst profiles based on the use of the physical CINR lookup table. A large number of lookup tables will be required as the mapping of burst profiles to CINR is expected to change with the channel type.

(b) FFR Management Department The role of the FFR management department is to maximize the spectral efficiency of the system. In general, as shown in FIG. 11, the effective spectral efficiency of the system is expected to change with the effective reuse factor.

The OFDMA frame of an IEEE 802.16 system usually starts from the reuse 3 zone. This is because the reuse 3 zone is defined as the transmission mode for frame preamble (see Figures 3 and 10). The FFR management department selects the optimum R3 / R1 (reuse 3 / reuse 1) transition CINR used by the zone allocation function in the station management department, and transmits it to the OTA (over-the-air) throughput (that is, to the PHY layer). The spectral efficiency is maximized by maximizing the number of bits of the payload.

If the threshold is too low, all users will be in the reuse zone 1. Interference occurs in some users around the sector edge, reducing CINR. This sets the burst profile very robustly, resulting in reduced throughput. Conversely, if the threshold is too high, all users will be in the reuse 3 zone. This guarantees the best throughput at the link level for all users, but the number of slots available for allocation is one-third of the number of slots available in the former case. Therefore, the throughput of the system is compromised. Therefore, by monitoring the effect of adjusting the threshold of OTA and using an algorithm based on a simple slope descent, the FFR management unit can optimally adjust the transition point.

(c) Connection scheduler To ensure that QoS requirements are met, the connection scheduler starts by considering the connections assigned to the UGS service type, and then the connections assigned to the real-time, enhanced real-time, and non-real-time service types. Proceed to, and finally to the best effort service type.

For each service type, the connection scheduler must determine whether a connection should be scheduled within a frame by analyzing the current QoS metrics for the QoS parameter set. .. The algorithm that determines this for each service type is considered below. This description first focuses on the downlink connection scheduler and emphasizes the differences from the uplink connection scheduler.

If the connection scheduler does not need to schedule a connection within a frame, but determines that the connection can be granted bandwidth if available resources are available, until the requirements for the remaining connections are considered. , The connection is put on hold. Details on how to determine if the connection should be considered after all others are described below.

It should be noted that for all DL connections, the SDU Buffer Size of the connection is zero or the connection is sleeping and the current frame is not the listening window, so the connection. Is not scheduled. Similarly, in UL, a connection is not scheduled if the Requested BW is zero and the connection is not issued an unsolicited grant or poll, or the same sleep condition described for DL exists.

For competing CDMA bandwidth requests received over an ERT, NRT or BE connection, resource reservation and insertion of CDMA_Allocation_IE into the UL MAP are resolved separately from the connection scheduler. Resource allocation for CDMA_Allocation_IE is done before the burst map section allocates resources to the scheduled connection. It should be noted that the maximum number of connections per frame is also set according to the resource limit of the MAC layer.

The connection scheduler processes the connections of various service classes outlined at the beginning as follows.

(c1) UGS Connections assigned to the UGS service type are set with the following QoS parameters associated with deciding whether to schedule the connection. -Tolerated jitter -Maximum latency -Request / Transmission policy -Unsolicited grant interval The connection scheduler exceeds the Maximum Allowed SDU Latency if the Next Grant Time is within a subframe (or has expired) or if the connection is not scheduled within a frame. Find out if it is. If the result is positive, scheduling the connection within that frame is considered. If none of these events occur, the connection is not scheduled within the current frame.

When all UGS connections have been reviewed, the connections scheduled for the current frame are ranked based on the Latest Grant Time shown below. Latest Grant Time = min (Next Grant Interval + Tolerated jitter, Next Grant Interval + Max.Allowed SDU Latency-Current Max.SDU Latency) The ranking is in ascending order of the last grant time.

For UL, the SDU latency is obtained via the FL field in the CID state. If FL is equal to or greater than Max Allowed SDU Latency, the connection is scheduled within that frame.

(c2) Real-time (RT) service The connection assigned to the RT service type is set with the following QoS parameters associated with deciding whether to schedule the connection. -Maximum latency -Minimum reserved traffic rate The connection scheduler checks if the SDU Latency is exceeded (that is, Maximum Allowed SDU Latency-Current Maximum SDU Latency &lt;frame period) if the connection is not scheduled within that frame. If the result is positive, scheduling the connection within that frame is considered. Also, if the connection is not scheduled within that frame, the current minimum rate is calculated using the following equation to see if the minimum guaranteed traffic rate is met. Current Minimum Rate = Last Grant Size / (Next Frame Time-Last Grant Time) If Current Minimum Rate &lt;Minimum reserved traffic rate, the connection is scheduled. If neither condition is met, the connection is revisited after the remaining class of service has been reviewed. In the case of UL, BS must provide regular polling opportunities based on the Unsolicited Polling Interval. To accommodate this, if the Next Poll Time is scheduled (or expired) within that frame, the Connection Scheduler schedules the connection. BS, as in UL, must ensure that the minimum reserved traffic rate is met. Therefore, even in UL, Current If Minimum Rate &lt;Minimum reserved traffic rate, the connection is scheduled.

(c3) Enhanced Real-Time (ERT) Service The connection assigned to the ERT service type is set with the following QoS parameters associated with deciding whether to schedule the connection. -Maximum latency -Tolerated jitter -Minimum reserved traffic rate -Unsolicited grant interval The same mechanism as described above for UGS and RT is used to determine whether to schedule connections based on allowed intervals, latency and minimum rates. In addition, the selected connections are ranked based on the criteria used for UGS. If none of the conditions are met, the connection will not be scheduled within that frame.

In the case of UL, the connection scheduler uses the same mechanism as described above for UGS and RT to determine whether to schedule connections based on intervals or rates. It should be noted that in the case of UL, ERT does not provide polling opportunities and always gives permission. Permissions are used to request additional bandwidth.

(c4) Non-real-time (NRT) service Connections assigned to the NRT service type are set with the following QoS parameters associated with deciding whether to schedule the connection. -Minimum reserved traffic rate The same mechanism as for RT is used to determine whether the connection should be scheduled based on the lowest rate guarantee. If the conditions are not met, the connection will be revisited after the remaining class of service has been reviewed. In UL, the connection scheduler provides polling opportunities, but the polling rate is around 1 second. Therefore, in the case of UL, the connection is scheduled on demand, ensuring that the minimum rate requirement or the next polling time is met (or expired).

(c5) Best effort (BE) service The connections assigned to BE and the connections assigned to RT and NRT that were not scheduled to meet the QoS requirements are considered last. Since the aforementioned mechanism results in allocating resources to meet the lowest QoS requirements, the remaining resources are allocated to maximize spectral efficiency while at the same time maintaining some fairness in the allocation of parts of the remaining resources.

Before doing this, the connection scheduler adds to the list of remaining connections all RT and NRT connections scheduled by the mechanism outlined above in relation to the class of service. This is to support the operation of the PDU formation unit, which initially allocates the minimum resources required to guarantee QoS. When this is done, the connection scheduler allocates the remaining resources to maximize spectral efficiency. Therefore, RT and NRT connections that have already been assigned a minimum value can be revisited to increase the allow size in the same way that unscheduled connections are considered.

The list of connections is then stored, for example, using equally fair criteria. The first step is to calculate the connection weights for each connection in list i. Wi = Last recorded CINR / (Averaged CINR) Next, the connections are stored in descending order of Wi. To ensure fairness is maintained when the weights of many connections are within the same range, the connection scheduler optionally lists these connections based on the Last Grant Time. Sort. UL uses the same mechanism to decide whether to schedule the remaining connections.

The mechanism for sorting unscheduled connections described above may be replaced by a system operator (or the criteria for that mechanism may be modified) to suit a particular application.

(c6) Other tasks As each service type is considered, a list of Connection Descriptors is built. This is an ordered list, starting with the top-level UGS connection descriptors, followed by the RT, ERT, and NRT connection descriptors in strict order. Next, the stored BE connection list is added to the connection descriptor list to form a complete list of connection descriptors supplied to the subframe forming unit. By the way, in some systems, it is desirable to prioritize ERT over RT in an ordered list.

The Connection Descriptors have the following information. -Pointer to CID Status (so parameters can be accessed) -Scheduling Reasons: Grant Interval, Latency, Min Rate, None, (Polling at UL).

The algorithm depends on the Service Type in the QoS parameter set and the reason for scheduling, so the reason for scheduling is to determine the method used to calculate the Grant Size for that CID. Used by the burst map section (more details will be explained in the "burst map section" section below). Reason "None" indicates that there is no minimum QoS requirement that results in scheduling of connections within a frame.

(d) Burst map section The burst map unit searches the list of connection descriptors supplied by the connection scheduler in order. In order to reach the allocated PDU size and update the MAP IE (or generate it in the case of UL), the burst map unit executes the following algorithm for each connection. 1 Position the target burst for the connection (based on a search within the zone assigned for the burst with the assigned burst profile). (Note: The limit of 256 connections per burst must be adhered to.) 2 If a burst is found, determine the maximum remaining resources in the slot. 3 Calculate the maximum allowed PDU allocation based on the available slots and the assigned burst profile format. Four Use the appropriate algorithm to determine the actual PDU size based on the class of service (as described below). It should be noted that the allocated size never exceeds the minimum amount of SDU remaining in the SDU buffer or the amount of resources available in the target burst. In the case of UL, it does not exceed the amount of BW requested for non-UGS connections (note that if the allocation is for polling, no BW requested is prominent, so The allocation is to be 6 bytes). Also, the burst size must take into account the request / transmission policy set for the connection as it can prevent fragmentation and / or packing. In this case, the burst map section must determine the size of the next SDU and use it to limit the allocation size.

Once the Grant Size is determined for each connection, the MAP message will be updated appropriately. In DL, the IE that describes the burst is updated to include the CID of the connection. In UL, IE is inserted into UL MAP if this connection is the first scheduled connection. If this connection is an additional scheduled connection, the allocation size recorded in IE is incremented by Grant Size (since there is only one IE per MS in UL). If none of the above termination conditions are met, the burst map section terminates processing of the list of Connection Descriptors. More specifically, the burst map section handles connections of different service classes as follows.

(d1) UGS UGS connections are scheduled because they are likely to exceed their allowed interval or SDU latency. For the former reason, the Grant Size can be calculated from the Maximum sustained traffic rate (equal to the minimum value) and the Unsolicited grant interval. Grant Size = Traffic Rate * Unsolicited grant interval If the connection is scheduled for the latter reason, the Grant Size is the amount of data contained in the SDU that is likely to exceed the latency requirement (that is, the number of bytes in the SDU that will experience the maximum delay). of Bytes in SDUs Experiencing Maximum Equal to the sum of the subheaders required to accommodate the size and packing of the Delay) (the value of the Number of SDUs Experiencing Maximum Delay if multiple SDUs are scheduled) ..

If the selected burst cannot accommodate this size of allowance, the burst map section attempts to locate another suitable burst in which the allocation can be made. If no burst is found, the Grant Size is adjusted to match the largest maximum allowed PDU size found in all possible target bursts. For UL, the same method is used. However, if the SI bit is set within the CID Status, the BS overallocates it by allocating 101% of the Grant Size.

(d2) Real-time (RT) service RT connections are scheduled because their SDU latency is likely to be exceeded or the minimum guaranteed rate is not met. For the former reason, the Grant Size is calculated as described above. For the latter reason, it is necessary to ensure that the size meets the minimum and maximum rate requirements. Therefore, permits are separated by: Minimum reserved traffic rate * (current time-Last Grant Time) &lt;Grant Size &lt;Max Sustained Traffic Rate * (current time-Last Grant Time) First, in order to maintain the minimum QoS requirements, the Grant Size is set at this stage, and the minimum guaranteed traffic rate (Minimum reserved traffic). Guarantee that the requirements of rate) are met. The connection is then revisited after the remaining service types have been considered for further bandwidth allocation. In UL, connections are scheduled because the Minimum reserved traffic rate is not met or the Unsolicited Polling Interval ends. In the former case, the same method is used to calculate the allowed size. In the latter case, the Grant Size is set equal to 6 bytes (to allow the BW Request Header to be sent). It should be noted that additional BWs may be assigned in the case of BE services by the mechanism described below.

(d3) Enhanced Real-Time (ERT) Service ERT connections are scheduled because their allowed interval, SDU latency is likely to be exceeded, or the minimum guaranteed rate is not met. In either of the first two cases, the mechanism described for UGS is used to calculate the allow size, in the other case the mechanism for rates for RT is used.

In the case of UL, ERT connections are scheduled according to allow interval or minimum rate requirements. In the case of UL, the appropriate mechanisms already described for UGS and RT are used.

(d4) Non-real-time (NRT) service NRT connections are scheduled because the minimum guaranteed rate is not met unless scheduled. Therefore, the Grant Size is calculated by: Grant Size = Minimum reserved traffic rate * (Current time-Last Grant Time) UL schedules according to the minimum guaranteed rate requirement, in which case the same mechanism is used to determine the Grant Size. Alternatively, if scheduled due to polling interval requirements, 6 bytes will be allocated for RT for the reasons mentioned above.

(d5) Best effort (BE) service and "no reason" The last BE connection and RT and NRT connections in the queue are scheduled because there is no data reserved and any QoS requirements have already been met. The burst map section searches for the remaining connections and attempts to make the maximum permission assignment for that connection if it has not yet been granted (or only polling allocation in the case of UL). If authorization (or polling assignments only in the case of UL) has already been done (which can happen for RT and NRT connections), the burst map section will maximize the size of those connection allocations as much as possible. Try to increase. In the case of RT and NRT, the maximum possible permit is determined by the Maximum sustained traffic rate as follows. Grant Size = Maximum sustained traffic rate * (Current time-Last Grant Time) In the case of BE, if this parameter is not set, the burst map section uses the maximum PDU size possible in consideration of the limits described above for the foreign car zone.

&lt;Summary&gt; In a detailed description of the "major" components, the functional and data structures that are the components, the interactions between the "major" components and between other components that are not considered part of the scheduler or PDU formation. Described the high-level architecture of the system of the low-level MAC scheduler and PDU formation part of the WiMAX base station.

The subsystem design ensures that the QoS requirements set for the active connection are met and that the spectral efficiency of the entire system is maximized through a combination of self-optimized partial frequency reuse and proportional fair scheduling. .. In the above embodiment, this is done by ensuring that each station is properly allocated to the most suitable reuse zone, and by allocating radio resources to the connection to ensure that the minimum QoS is met first. , Achieved. The remaining radio resources are then allocated on a proportional fair basis. However, proportional fairing is not an algorithm that is only possible to allocate the remaining resources. Another possible approach is to allocate based solely on the CINR of each connection. In this case, "fairness" is provided at the QoS level, for example when assigning a class of service to a connection.

Subsystems appear to be somewhat complex, but this is specific to solutions that ensure that they meet the minimum QoS requirements. However, the design of all component functions and algorithms is simple and well-structured and aims to limit the amount of processing power required to implement the subsystem.

The above description refers to, by way of example, a SISO radio communication system, a system in which a single BS antenna exchanges signals with a single SS antenna. However, MIMO systems can also use the subsystems of the present invention. In MIMO systems, there are multiple antennas (and associated receiver circuits) at each end of the connection.

The embodiments of the present invention may be implemented as hardware, or software modules that operate on one or more processors or a combination thereof. That is, one of ordinary skill in the art will appreciate that a microprocessor or digital signal processor (DSP) may actually be used to perform some or all of the functions of the subsystem described above. The present invention may also be implemented as a program of one or more elements or devices (eg, computer programs and computer program products) that performs some or all of the methods described herein. Such a program carrying out the present invention may be stored on a computer-readable medium or may be in the form of, for example, one or more signals. Such a signal may be a data signal of any other form that can be downloaded from an internet website or provided by a carrier signal.

The present invention relates to a wireless communication system in which a base station communicates with a subscriber station by exchanging data packets contained within a frame. The frame occupies a specific time period and frequency range. Each frame has a downlink subframe from the base station to the subscriber station and an uplink subframe used by the subscriber to transmit data to the base station. Each subframe has multiple zones, and each base station has multiple simultaneous connections with subscribers by transmitting packets to the associated subscriber stations within the zone. Zones allow frequency reuse when subscribers are located in different cells served by the same base station. In addition, each connection has an associated service level that more or less demands resources in the system.

A base station that implements the present invention has a subsystem that provides packet scheduling and partial frequency reuse management. The base station is: a means of receiving packets sent from the subscriber station over each connection; a time-critical connection that must first be scheduled within the current downlink subframe to achieve service levels. Then determine the non-time-critical connections that can be scheduled within the subframe using the resources that are still available, and then assign the priority of the non-time-critical connections within the subframe. Means for packet scheduling; and subframes of available frequencies available for transmission from the base station by assigning zones to each connection based on a comparison between the signal level from the subscriber station and the variable threshold. Have means to perform reuse;

The priority given to non-time-critical connections during scheduling may be governed by parameters set by the system operator. On the other hand, the service level used to determine time-critical connections is set to system standards. The thresholds used for partial frequency reuse are selected to provide optimal reuse 3 / reuse 1 transitions for subscribers within a particular cell.

Although the IEEE 802.16 wireless communication system has been described above as an example, the present invention may be applied to other communication systems that require packet scheduling at the MAC layer of the protocol stack.