Admitting data flows to a multiple access network

ES 2 392 949 T3

DESCRIPTION

Support of data streams to a multiple access network

Field of the invention

The present invention relates generally to communication systems and networks and more specifically to multiple access networks.

Background

Multiple access networks allow network operators to exploit the statistical multiplexing gain that arises from a number of users sharing network resources. However, the excessive allocation of resources can lead to adverse operating conditions, such as unacceptably long delays. Provision of guaranteed quality of service (QoS) levels at higher priority data flows than best effort data flows attempts to solve this problem. However, without admission control to avoid a high priority class overload, CdS guarantees cannot be given or maintained. Thus, prioritized channel access must be coupled with admission control which limits the number of high priority flows.

Wireless Local Area Network (WLAN) technologies have achieved enormous growth in popularity and are currently being deployed in homes, offices and public access spaces, particularly as the last hop network. The growing popularity of streaming media and Voice over Internet Protocol (VoIP) applications requires such WLANs to incorporate quality of service (QoS) support. CdS support is currently being standardized in the IEEE 802.11 standard for wireless networks. Two mechanisms will be supported, namely a simple priority-based approach and a more complex and centralized scheduler-based approach to use where tight control of oscillatory disturbances is required. The prioritized access to the network is implemented using differentiated Media Access Protocol (MSC) variables for the different priority flows, such that the high priority flows have access to the network in preference to the low priority flows. Wide deployment of this simple mechanism is anticipated.

Admission control has been widely studied in the context of wide area networks such as Asynchronous Transfer Mode (ATM) and Internet Protocol (IP) networks. However, such networks operate at high speeds over reliable wired or fiber optic links. At high transmission speeds, admission control can be performed in a suboptimal way by allocating resources conservatively and reserving more bandwidth than is actually necessary to achieve satisfactory CdS levels.

On the other hand, wireless networks have low transmission rates and dynamic channel characteristics. Data reliability is generally of greater importance than delay, and robust transmission link methods are needed, such as return channelless error correction and automatic repeat request. Consequently, the results of studies relating to high-speed data networks cannot be easily applied to WLANs or other networks exhibiting characteristics similar to WLANs.

It is known in the art to employ an admission control function that uses a measurement-based approach to calculate the load on the network. The function makes intake control decisions based on the calculated load. Admission control relies on the presence of a timestamp in each data packet to estimate delays. However, a timestamp is not available in a WLAN unless it is inserted by an application.

US Patent No. 6,216,006 describes a method of allowing new traffic on a wireless data network. A service request is received from a mobile station. The effective bandwidth of the service request is estimated using a recursive estimator. If there is enough excess capacity in the wireless data network to accommodate the service request, the service request is granted. In response, a discrete fit, based on effective bandwidth, is applied to the recursive estimator. The recursive estimator is used to estimate the current used / unused bandwidth. Average bit rate is added / subtracted as a correction whenever a call is admitted or terminated. The bandwidth requirement for new calls is estimated as a weighted sum of average and maximum bit rates. Calls are allowed if the bandwidth requirement of the new call is less than the estimated unused bandwidth.

Summary

According to the present invention, a method is provided for supporting a data stream corresponding to a new service request to a multiple access network, a network access point for supporting a data stream to a multiple access data network, and a communication system as set forth in the accompanying claims.

Aspects of the present invention provide a method, a network access point and a communication system for admitting data streams corresponding to new service requests to a multiple access network. The method includes the steps of receiving from a network node a service request that has quality of service requirements.

ES 2 392 949 T3 service (CdS), calculate the effective bandwidth required by a data flow corresponding to the service request based on an average packet size of the data flow, determine the maximum capacity of the network, and admit the data flow corresponding to the service request if the network has sufficient capacity to accommodate the effective bandwidth required by the data flow.

The average packet size of the data stream is preferably determined from the service request, and the effective bandwidth required by the data stream corresponding to the service request can be optionally adjusted if the transmission rate of the request service is less than the maximum transmission rate of the network. Preferably, if the number of transmitting nodes increases due to the new data flow, the maximum capacity of the network is determined based on the number of transmitting nodes in the network.

In a preferred embodiment of the invention, the data flow corresponding to the service request is admitted to the network if the sum of the bandwidth occupied in the network and the effective bandwidth required by the data flow is less or equal to the maximum capacity of the network.

It is also preferable to store and / or update the number of transmitting nodes and the occupied bandwidth of the network for future use when supporting other data streams. Likewise, the number of transmitting nodes and the occupied bandwidth of the network can be stored and / or updated when a data flow ends.

Another aspect of the present invention provides a network access point to support a data stream to a multiple access data network. The apparatus includes a receiver to receive from a network node a service request that has quality of service (QoS) requirements, at least one processor to calculate an effective bandwidth required by the data flow corresponding to the service request. based on an average packet size of the data stream and to determine a network capacity based on a number of transmitting nodes in the network, an admission controller to admit the data flow relating to the service request to the network provided that the network has sufficient capacity to accommodate the effective bandwidth required by the data flow and a transmitter to transmit the admission grant to the network node.

Another aspect of the present invention provides a communication system, including a multiple access data network, a plurality of network nodes capable of communicating via the network, and at least one access point coupled to the multiple access data network. An access point includes a receiver to receive from a network node a service request that has quality of service (QoS) requirements, at least one processor to calculate an effective bandwidth required by a data flow corresponding to the request. based on an average packet size of the data stream and to determine a network capacity based on a number of transmitting nodes in the network, an admission controller to admit the data flow related to the service request to the network provided that the network has sufficient capacity to accommodate the effective bandwidth required by the data flow and a transmitter to transmit the admission grant to the network node.

Brief description of the drawings

Embodiments of the present invention are described below, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 is a graph depicting typical performance characteristics of a WLAN.

Figure 2 is a graph representing the effect of data packet size on WLAN capacity.

Figure 3 is a graph representing the effect of the number of transmitter nodes on the WLAN capacity.

Figure 4 is a flow chart of a method for admission control of new traffic in a WLAN according to an embodiment of the present invention.

Figure 5 is a flow diagram of a method for admission control of new traffic in a WLAN according to an embodiment of the present invention.

And Figure 6 is a block diagram of an architecture for implementation of the methods described below with reference to Figures 4 and 5.

Detailed description

For ease of explanation, a method, apparatus, and system are described below with specific reference to a Wireless Local Area Network (WLAN). However, the embodiments of the present invention are not intended to be limited thereto since the principles of the method and apparatus described below have general applicability to other types of networks exhibiting characteristics similar to WLANs. For example, embodiments of the invention are also applicable to Phoneline Networking Alliance (HPNA) home networks, which include cable networks that possess similar performance characteristics to WLANs. Specifically, HPNA networks

ES 2 392 949 T3 have a prioritized MAC, multiple rates and a long preamble. Another example is the HomePlug Powerline Alliance (HomePlug), which currently does not have CdS support, but will likely be standardized on a similar CdS mechanism in the future. The method, apparatus and system described below have application to multi-class multiple access LANs incorporating CdS support.

A typical performance characteristic of a multiple access WLAN is depicted in Figure 1, which is a graph of the mean delay as a function of offered load. As the offered load progressively increases, the mean delay gradually increases until a threshold 110 is reached (indicated by a vertical dotted line), after which the delay increases sharply (substantially vertically). Threshold 110 is called the capacity of the network. Provided that the high priority traffic is limited to within the capacity of the network, the CdS characteristics of the network are satisfactory or good, and the network is considered to offer a controlled load service. A prioritized WLAN can thus prevent best-effort traffic from affecting high-priority flows. Other CdS characteristics, such as jitter, are affected similarly to mean delay. The purpose of admission control is to guarantee CdS by ensuring that the offered load remains below the capacity of the network. However, the capacity of the network is not constant, and the admission control algorithm and thus the apparatus has to take into account the variations in capacity. Network capacity is affected by the following factors:

* Mixing packets on the network. WLAN packets typically have long preambles due to the need to interoperate with nodes at different data rates. For small packets, the size of the preamble is of the same order of magnitude as the size of the actual data packets. Voice traffic has small packet sizes and consequently reduces network capacity. On the other hand, video traffic has large packet sizes and thus increases the capacity of the network. The capacity of the network therefore depends on the average packet size or the mix of packets in a WLAN. Figure 2 is a graph representing WLAN capacity as a function of packet size. Network capacity increases non-linearly as packet size increases, largely semi-parabolic, until a maximum packet size 210 is reached. As can be seen, network capacity increases rapidly as a function of packet size with smaller packet sizes, but increases less rapidly as the average packet size increases and approaches the Maximum Transmission Unit (MTU).

* The number of transmitting nodes in the network. The capacity of a WLAN is also affected by the number of transmitting nodes, due to an increase in the number of collisions when the number of transmitting nodes increases. Figure 3 is a graph of network capacity as a function of the number of transmitting nodes. As can be seen from the graph, the capacity of the network is reduced when a constant load is offered by an increasing number of nodes. However, the graph is substantially flatter than Figure 2, indicating that the number of transmitting nodes does not have as great an effect on network capacity as the effect of average packet size.

* The physical transmission rate in a multi-rate WLAN. WLANs support transmissions at various physical rates. Nodes transmit at different rates depending on factors such as link quality and distance from the access point (AP). Transmission at a rate lower than the maximum rate reduces the overall capacity of the network as this results in the channel being busy for a longer period of time. Thus, when a new data flow is requested, the current values ​​of the factors affecting the capacity of the network are obtained and compensated. Compensation can be done by reducing network capacity, calculating an effective bandwidth for the new flow, or a combination thereof. A new data stream is only allowed if the network has sufficient capacity to accommodate the effective bandwidth required by the data stream.

Figure 4 represents a flow chart of a method for admission control of new traffic in a WLAN.

In step 410, a service request for a new data stream is received from a WLAN node. The effective bandwidth for the new data stream is calculated in step 420. The capacity of the network is determined in step 430. In step 440, it is determined whether the network has sufficient capacity to accommodate the effective bandwidth. required by the new data flow. If so (YES), the new data flow (service request) is admitted in step 445 and the process ends. If the new data stream cannot be accommodated due to insufficient network capacity (NO), the process ends directly.

Figure 5 depicts a more detailed flow diagram of a method for admission control of new traffic in a WLAN.

In step 510, a service request for a new data stream is received from a WLAN node. The requested bandwidth with respect to the new data flow is determined from the CdS signaling between the requesting station and the WLAN admission controller, in step 520. In step 530, the effective bandwidth of the new data flow Data is calculated using the average packet size of the new data stream. Average packet size is obtained directly from CdS signaling.

The effective bandwidth BWf of the new data stream to compensate for the mix of packets in the new

ES 2 392 949 T3 data stream is:

Bw_andff ~ (YES6p_ac]_{andt + SÍZCheader) / SlZCpacket * BWactual where:

BWactuai is the requested bandwidth,

Sizepacket is the average package size, and

Sizeheader is the WLAN header size, which is a known constant for a specific network.

In step 540, it is determined whether the new data flow increases the number of transmitter nodes. The number of transmitting nodes can be determined using CdS signaling, which indicates the address of a transmitting node. If the number of transmitter nodes increases when the new data stream is added (YES), the new maximum network capacity is determined in step 545, before processing continues in step 550. The maximum capacity of the network as a function of the number of transmitting nodes is determined in advance by measurement or simulation. The relevant value for the maximum capacity is typically obtained from a look-up table, based or indexed on the number of transmitting nodes. In the absence of an exact match in the lookup table, the next highest value in the table is selected. Table 1 provides an example of a look-up table to determine the maximum network capacity as a function of the number of transmitting nodes:

Table 1

<td>Number of TX nodes</td><td> 1</td><td> 2</td><td> 3</td>

<td>Maximum capacity (Mbps)</td><td> 11</td><td> 10,8</td><td> 10,6</td>

If the number of transmitter nodes did not increase (NO), processing continues from step 540 to step 550.

In step 550, the physical transmission rate of a data stream relative to the new service request is determined by means of a link supervisor at the WLAN access point. Nodes can periodically vary their transmission rate to cope with link conditions and station mobility. By having a link supervisor continuously sample packets from each node, variations in network capacity can be tracked dynamically taking into account changes in physical rate. The link supervisor obtains the physical rate from the WLAN headers.

In step 560, it is determined whether the physical rate is less than the maximum rate. If the physical transmission rate is less than the maximum rate determined in step 545 (YES), the effective bandwidth BWf of the new data stream is determined according to the following formula, in an adjustment or calculation step 565:

<img file="ES2392949T3_D0001.tif" />

where:

BWeff is the effective bandwidth at the maximum rate Rmax,

Rmax is the maximum rate, and

Ractual is the physical data rate.

From step 565, processing continues to step 570. If the physical rate is not less than the maximum rate (NO), processing continues from step 560 to step 570.

In step 570, it is determined whether the total bandwidth requirements of the already supported flows (occupied capacity) and the new flow (BWeff) is less than or equal to the maximum network capacity, which was determined in step 545. If so (YES), the new data stream (service request) is admitted in step 575 and the number of transmitting nodes and the new occupied bandwidth (capacity) are updated and stored in step 580. The new occupied capacity is equal to the sum of the previous occupied capacity and the effective bandwidth (BWeff) of the new data stream. Processing then ends after step 580. Otherwise, if step 570 is false (NO), processing ends.

The admission controller updates its status similar to when data flows end, about when new data flows are admitted to the network.

ES 2 392 949 T3

Figure 6 is a block diagram of an architecture with which the methods of Figures 4 and 5 can be implemented.

Referring to Figure 6, a WLAN Access Point (AP) 620 is connected to a wired Local Area Network (LAN) 610 and provides wireless access to LAN 610 for wireless nodes 660 and 665. Although not illustrated To simplify the drawing, those skilled in the art will appreciate that other devices may be connected to the LAN 610, as indicated by the branched lines. The WLAN AP 620 includes a Subnet Bandwidth Manager (SBM) 630, which includes a 675 processor or computer to run software programs to perform functions such as calculations and data storage and retrieval, and a transmitter and receiver (transceiver ) 685 connected to an antenna 690 (typically an omni-directional antenna) to communicate with wireless nodes 660 and 665. The transceiver 685 in use allows to receive the service request from a network node, such as nodes 660 and 665, with the service request having the quality of service (QoS) requirements. Transceiver 685 also provides a transmit grant of said admission to a network node such as nodes 660 and 665. The SBM 630 also includes a WLAN admission controller 640, a link supervisor 650 (a monitoring unit to determine a transmission rate of the service request), a memory unit 695, for example formed by a semiconductor memory of random access (RAM) and / or a read-only memory (ROM), and a 670 I / O interface to connect to the 610 wired LAN.

The 675 processor can be any processor, for example, it can be implemented using an IEEE 802.11 MAC processor or a wireless processor. Many other processors and processing units can be practiced without departing from the scope and spirit of the invention. The 685 processor allows you to: determine the average packet size; calculating the effective bandwidth required by the data stream corresponding to the service request based on an average packet size of the data stream; and determining the capacity of a multiple access data network in the WLAN 710. Furthermore, the processor 675 determines the capacity of the WLAN 710 based on a number of transmitter nodes in the WLAN 710 if the number of transmitter nodes increases due to data flow. In addition, the admission controller 640, in use, allows the WLAN 710 to support the data stream relative to the service request provided that the WLAN 710 has sufficient capacity to accommodate the effective bandwidth required by the data stream. . Components 640, 650, 670, 675, 685, and 695 of access point 620 communicate via an interconnected bus 700 and in a manner that results in a mode of operation known to those skilled in the relevant art.

The wireless nodes 660 and 665 send the flow characteristics and CdS requirements to the SBM 630 using the ReSerVation protocol (RSVP). RSVP is typically used to request network-specific qualities of service for particular application data streams. An RSVP message contains a traffic specification, which specifies the bandwidth requirements as well as the average packet size to be used by the CdS flow. The RSVP message also contains the address of the sending node. Although other methods or protocols can be used for this purpose, RSVP is a standardized method to perform such signaling.

The method, network access point and communication system described above provide a level of CdS to applications in multiple access networks that correspond to a level of service comparable to that of a lightly loaded WLAN or a cable network that exhibits substantially characteristics similar to a WLAN. Thus, a prioritized WLAN or similar cable network incorporating an embodiment of the present invention can provide a light load impression to high priority traffic, even when overloaded with background best-effort traffic. Such a service is desirable and suitable to support VoIP and multimedia streaming applications, among other types of data streams. However, actual applications using the WLAN or similar wired network are assumed to provide enough buffering to cope with smaller variations in CdS.

The detailed description provides an exemplary preferred embodiment only, and is not intended to limit the scope, applicability, or configurations of the invention. Rather, the description of the exemplary preferred embodiment provides those skilled in the art with descriptions that enable them to implement the exemplary preferred embodiment of the invention. It should be understood that various changes can be made to the function and arrangement of the elements without departing from the scope of the invention set forth in the appended claims.