Improved wireless local area network

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Technical field to which the invention belongs]

The present invention relates to a wireless data communication network, and more particularly to a communication device that uses wireless data communication between a mobile data processing unit and a central computer.

【0002】

[Conventional technology]

Applicants of the present invention provide a wireless data communication system known as the Spectrum 24 system, which follows the IEEE Standard 802.11 wireless data communication protocol. When the system is implemented, the mobile unit communicates data with the central computer through the access point. The access point may communicate with one or more central computers over a wired network. Each mobile unit associates itself with one of the access points. The access points of this system are functioning to fulfill all the requirements of the standard protocol, including association and movement functions, packet configuration and parsing, packet splitting and reassembly coding, and the system. Includes access control. To maintain order and reduce wireless traffic, each access point determines which of the data communications received from the central computer over the wired network is destined for the mobile unit associated with that particular access point. There must be. This requirement adds considerable computational capacity to the access point and increases its cost.

【0003】

Moreover, self-service shopping system, hospital systems, systems that include paging or voice data links to many users, support communications with the electronic shelf labels such systems bets, needs to support a large amount of data communications from multiple users Some applications require additional access points to support data communication traffic, increasing the overall cost of the system. The cost of a working access point is not only its complexity and the need for fast processing of data packets to select packets for each mobile unit associated with the access point, but also the delivery of power to the location of the access point. It also depends on incidental costs such as the cost of converting alternating current to direct current for access point circuits. In addition, costs can also be associated with the physical installation of access point hardware and antennas.

【0004】

In a conventional system, each access point is connected to a central computer by an Ethernet wired network. The access point is required to determine the identity of the mobile unit associated with the access point and extract packets destined for the mobile unit associated with the access point from the data packets on the Ethernet network. This requirement has been a significant processing burden on the access point and has increased the cost of the access point. In the previous system described in International Patent Application WO 09937047 published by this applicant on 22 July 1999, the central computer communicates with a high-performance relay hub over an Ethernet wired network. Alternatively, a Token Ring network can be used. The relay hub determines the destination of each packet, and if the destination of the packet is a mobile unit related to the access point, routes the packet to the access point. To achieve this functionality, the hub must be a sophisticated hub that maintains a route list of mobile units and their associated access points that follow the hub's ports.

【0005】

In practice, the hub only needs to maintain an original list of access points connected to the hub and mobile units associated with the access points connected to the hub. If a packet is received on Ethernet to a hub with a destination address that is not associated with that hub, the packet is ignored. The hub will only forward the packet to the access point if the destination address of the packet is known by the list. When a packet is received on a hub port associated with a communication line connected to the access point, the source address is associated with the hub port in the list. The packet is either an Ethernet connection or routed to another port according to the destination address.

【0006】

By determining the destination address at the hub and maintaining the association between the hub port and the mobile unit address with the connected access point in the hub's routing list, the functionality required of the access point is greatly reduced. The access point acts only as a conduit that sends RF transmissions of packets received on its line of communication, receives transmissions from associated mobile units, and supplies Ethernet packets to the hub. In addition, the access point must provide mobile unit-related features as provided by the Spectrum 24 system and other 802.11 protocol features, and may provide surrogate polling responses for related mobile units in labor-saving mode. ..

【0007】

[Problems to be Solved by the Invention]

Traditional systems may have a large number of access points, each with memory containing program instructions that perform a variety of required functions. This distributed process makes it difficult to upgrade or make changes to the system configuration because any upgrade or change may require changes to the program code on each of the access points. Such distribution of processing unit functions also makes system management functions such as load balancing and access control more difficult. Therefore, an object of the present invention is a low cost improvement that enables the economical provision of reliable wireless data communication by increasing the capacity in a complex device, at a reasonable cost, or by a simple device. The purpose is to provide a type wireless data communication method and system.

【0008】

[Means for solving problems]

The present invention provides a system for wireless data communication between a mobile unit and a wired network. The system has at least one data interface and is arranged to receive the formatted data signal at the data interface and transmit the corresponding RF data signal, and receive the RF data signal and the corresponding formatted data signal. Includes multiple RF ports that are arranged to prepare. Also, at least one arranged to receive a data signal from the wired network, prepare a corresponding formatted data signal, receive the formatted data signal and supply the corresponding data signal to the wired network. A cell control device is provided, which controls the connection of the mobile unit with one of the RF ports, supplies a formatted data signal for the mobile unit to the associated RF port, and formats data from the mobile unit. Receive the signal from the associated RF port.

【0009】

The present invention provides improvements in a wireless data communication network with multiple RF ports and mobile units coupled to a data processing system, the mobile unit being tied to one of the RF data communication ports to process the data. Perform data communication with the system. The mobile unit is assigned to one of the RF ports by the cell controller, which receives the first data communication from the data processing system and relays the data communication to the assigned RF port. It is arranged so that the second data communication is received from the RF port and the second data communication is relayed to the data processing system.

【0010】

INDUSTRIAL APPLICABILITY The present invention provides a method of operating a radio local area network having at least one RF port, a plurality of mobile units, and a cell control device combined with the RF port. By operating the RF port, the received signal from the mobile unit is relayed to the cell control device, and the received signal from the cell control device is relayed to the mobile unit. The cell controller is operated to control the connection of the mobile unit to the RF port, which is the transmission and reception of the coupling signal between the RF port and the cell controller, and the mobile unit via the RF port. Including sending and receiving messages.

【0011】

INDUSTRIAL APPLICABILITY According to the present invention, a mobile unit used in a wireless data communication system is improved, and the unit has a data processor, a program for a data processor, and a wireless network adapter including a programmed processor and a wireless module. The programmed processor performs a first communication processor function, including control of the radio module, and the data processor operates under the program, with a second including association to the radio access location of the radio data communication system. Performs communication processor functions.

【0012】

The present invention provides improvements to wireless data communication systems to provide data communications according to standardized protocols, including associations with mobile unit radio access locations. The radio access location is provided with at least one RF port, including a radio module and RF port processor that communicates data with the programmed computer. The RF port processor performs the first function of the standardization protocol, and the programmed computer performs the second function of the standardization protocol, including the association of the mobile unit with said radio access location.

【0013】

According to the present invention, it is used in a wireless data communication system including a wireless module having a data interface and a transmitter / receiver for wireless data communication, and a digital signal processor having first and second data communication ports, random access memory and ROM. An RF port is provided. The second data communication port is coupled to the data interface of the wireless module. The ROM includes a bootstrap loader program to control the digital signal processor, loading program instructions into random access memory through the first communication port.

【0014】

The present invention provides a method of operating an RF port with a radio module, a digital processor, random access memory, and ROM. The bootstrap loader program is stored in ROM. The digital processor is activated to download instructions from the computer to random access memory using the bootstrap loader, and the RF port is activated to send and receive messages using the radio module under the downloaded instructions. Will be done.

【0015】

The present invention provides a method of transmitting a signal having a radio signal format using a wired network interface, a data processor, and an RF port having an RF module. The signal is fed to a wired network interface that has radio address data and message data in a data packet destined for the RF port using the protocol for the wired network. The processor is activated to supply a radio signal format for address data and a radio data signal with message data to the RF module so that the RF module operation transmits the radio data signal as an RF signal modulated in the radio signal format. Is activated.

【0016】

The present invention provides a method of transmitting a signal having a radio signal format using an Ethernet interface, a data processor, and an RF port having an RF module. An Ethernet data packet that encloses a data message having a wireless signal format as data is supplied to an Ethernet interface. The data processor is operated to deliver data messages to the RF module. The RF module is operated to send a data message as an RF signal.

【0017】

The present invention provides a method of receiving a signal having a radio signal format including radio address data and message data at a wired network interface, a data processor, and an RF port having an RF module. The RF module is operated to receive an RF signal with a radio signal format. The data processor is activated to receive the wireless data signal from the RF module and use the wired network protocol to deliver the data signal to the wired network interface containing the data packet with the source address corresponding to the RF port. Here, the data packet includes radio address data and message data.

【0018】

The present invention provides a method of receiving an RF message signal having a radio signal format including an address data format and message data using an Ethernet interface, a data processor, and an RF port having an RF module. The RF message signal is received by the RF module and supplied to the data processor as a data signal. The data processor is operated to translate the address data into a data signal, and based on the address data, the message data and the address data are enclosed in the Ethernet data and supplied to the Ethernet interface.

【0019】

The present invention provides a simplified wireless local area network system that includes a computer with a data processor and memory, an RF port with an RF port data processor, an RF module, and a data communication interface coupled to the computer. The first program is prepared in the memory of the computer to operate the computer data processor and performs the first wireless data communication function including the association to each mobile unit. The second program is prepared to operate the RF port data processor and performs the second radio data communication function.

【0020】

According to the present invention, a wireless access device is provided to prepare wireless access in a communication system. The device includes a modem for sending and receiving data messages on a communication system, and an RF port having a data interface coupled to the modem, a data processor, and an RF module. The data is programmed to receive data messages from the modem, format the messages for wireless data communication, and feed the formatted messages to the RF module for transmission by RF data signals to at least one remote station. , It is possible to receive RF data signals from at least one of its remote stations and supply the data messages to the modem for transmission to the communication system.

【0021】

The present invention provides a method of providing wireless access to Ethernet. A modem with a data communication interface coupled to the RF port is connected to the Internet. The RF port is formed to perform radio data communication with at least one mobile unit having a predetermined radio communication address. A mobile unit formed to have a predetermined radio communication address is prepared for RF data communication with an RF port. The RF port is arranged to relay communication between the mobile unit and the modem.

【0022】

The devices and methods of the present invention provide RF ports for wireless access locations, which are less expensive than known access points and offer greater system manageability and flexibility. Most of the software used to control communication with the mobile unit runs on the controller, where software upgrades and changes are easily carried out. In some embodiments, the directives are downloaded to the RF port, making it easier to upgrade the RF port directives. System control is centralized, easy to manage, and allows access control and coding function changes. Priority for traffic can also be established to facilitate digital telephone communication by prioritizing voice traffic. Therefore, a system with considerable flexibility is provided using common RF port hardware that provides a wireless LAN with one to hundreds of wireless access locations. In order to better understand the present invention, the following description will be referred to with the accompanying drawings along with other and further embodiments of the present invention, the scope of which is pointed out in the scope of the attachment.

【0023】

BEST MODE FOR CARRYING OUT THE INVENTION

Referring to FIG. 1, an example of a wireless data communication system 10 according to the invention that provides data communication between a central computer or group of computers on a wired network 16 and a plurality of mobile units 20 is shown. Whereas traditional systems used access points that could manage wireless communication with mobile units at each wireless access location, the system in Figure 1 used a simplified RF port 18 at each wireless access location. Uses wireless communication protocols such as IEEE standard 802.11 to provide wireless packet communication with the mobile unit 20, thereby allowing the wireless module within the mobile unit 20 to derive from the cell controller 14 and be an RF port for data communication purposes. Monitor the poll signal from RF port 18 associated with 18. The system configuration of Figure 1 is particularly effective for large wireless local area networks (LANs) where a large number of wireless access locations need to be prepared. Such systems typically operate at low power microwave frequencies and require radio access locations approximately every 100 feet. Hundreds of wireless LAN systems located across large facilities such as enterprises, hospital facilities, university campuses, where they need to be operated in conjunction with mobile units such as portable computers and similar devices. A large number of such wireless access locations are needed. Therefore, there is an incentive to reduce the installation cost at each wireless access location. According to the present invention, the system configuration and operation are redesigned to reduce the cost of each of the individual wireless access points. Further, the system of the present invention concentrates operational control on one or more central control units 14, facilitates system management, and facilitates installation of modifications and upgrades.

【0024】

According to the present invention, much of the functionality of the 802.11 protocol associated with conventional access points is removed from the device attached to the wireless access location and provided in the cell control device 14, where the cell control device 14 is a wireless network. It may be installed with a relay hub 12 that connects 10 to the associated wired network 16. In particular, a conventional "access point" device has been replaced by a simplified device 18, referred to herein as an "RF port" that includes an RF module, which is used in prior art access points. It may be the same RF port as the one, and is a simplified digital circuit that performs only a limited part of the 802.11 medium access control (MAC) function performed by the prior art access point. In particular, RF port 18 preferably requires only a low level of processing power in terms of processor capacity and software complexity (memory requirements) and performs only time-sensitive access point functions. Other functions that require more processor-intensive and complex programs and are less time-sensitive are delegated to one or more "cell controllers" 14, which are more complex for multiple RF ports 18. Perform various functions. The term "RF port" is intended to be construed as including a circuit with an antenna. In particular, it is intended to represent any means of connecting a radio airway to a wired medium and / or vice versa. More specifically, the port can be, at a minimum, an antenna and an International Organization for Standardization (ISO) physical layered circuit (ie, an RF modulator / demodulator).

【0025】

In order to perform the high level processing unit function of the access point in the cell controller 14, according to the present invention, all messages to or from the mobile unit 20 associated with a particular RF port 18 are sent to the cell controller. Processed within 14. The system has one or more cell controls, each of which includes a level computer with, for example, a Pentium® type board, each of which has data message traffic and multiple RF ports 18 selected. Deployed and programmed to handle mobile unit associations with. A relay hub 12 is inserted in between, and a message for switching between the communication line 16, the RF port 18, and the wired network connected to the cell control device 14 is prepared. Each of the one or more cell controllers 14 acts as a virtual "access point" for traffic destined for the RF port 18 with which the traffic is associated and the mobile unit 20 associated with those RF ports. When the message addressed to the mobile unit 20 is received on the communication line 16, the relay hub 12 directs the message to the appropriate cell controller 14, which reformulates the message and re-formats the message through the relay hub 12. To relay to RF port 18. When the message is received by RF port 18, it is converted to a radio message with minimal processing and sent to the mobile unit 20. Similarly, when a message from the mobile unit 20 is received by RF port 18, it is converted into a digital message packet and relayed through the relay hub 12 to the cell controller 14 associated with RF port 18. The cell controller 14 parses the message for further relay in the system.

【0026】

An important feature of the preferred embodiment of the present invention is the fact that the mobile unit association with the RF port 18 is a function handled by the cell control device 14. Therefore, when the mobile unit 20 is initially activated, it sends a related request signal in response to the beacon signal transmitted by the RF port 18 (in response to a command from the cell controller). The related request signal is relayed by the RF port 18 to the cell controller 14, which performs the necessary processing related to the RF port, including the possibility of loading. The cell control device 14 generates an appropriate response signal transmitted by the RF port 18 to the mobile unit 20. The cell controller 14 is in an appropriate position to evaluate the load on RF port 18 under its control, and thus easily load equalizes, for example, by feeding the RF port 18 a message of acceptance or rejection of the relevant request. Can perform the function. Further, the cell control device 14 receives a load message from another cell control device 14 in the system 10 and thereby adjusts the overall load management. Since the mobile unit 20 moves from the position where it receives service from one RF port 18 to the position where it receives service from another RF port 18, the cell controller 14 moves from the mobile unit 20 from various RF ports in the system. Receives information indicating the reception of the beacon signal of the mobile unit 20 and performs the functions necessary to support the movement of the mobile unit 20.

【0027】

In system 10 of FIG. 1, the cell controller 14 is shown as a separate computer connected to the relay hub 12, but the term "cell controller" is executed by these computers rather than the computers themselves. It is intended to refer to the logical function to be performed. As will be revealed below, the cell controller may be implemented in a variety of ways different from those shown in the exemplary system 10 of FIG. Simplified RF port implementation is done by performing the "high level" function of the cell controller's 802.11 protocol, Medium Access Control (MAC), and the simplified RF port "low level" function. Is achieved by executing.

【0028】

Low-level features are hardware-intensive features and are often time-sensitive. High-level features are software-intensive features and are not time-sensitive. One possible division of the exemplary 802.11 MAC function is: Low level functionality (preferably running on RF port) Cyclic Redundancy Check (CRC) Network activity vector (NAV) Ready to send / cancel send (RTS / CTS) Header generation / parsing Collision avoidance Frequency hopping Parsing Approval / Generation Resend time expired High level functionality (preferably performed on the cell controller) Related processing Move Retransmission Speed control Host interface [0029]

The following optional (high or low) level MAC features can be included in both high and low level classifications. Wired Equivalent Personal Encoding / Decoding (WEP) Split / reassemble Data movement Labor-saving polling support (PSP) According to the preferred system arrangement of the present invention, the low level MAC function is prepared by the RF port, the high level MAC function is prepared by the cell controller, and the optional level function is provided by either the cell controller or the RF port. Can also be prepared.

【0030】

The main advantage of the present invention is the cost savings of the processor capacity and storage capacity of the hardware or RF port. For example, a system with more than 100 wireless access locations can be implemented with one or two cell controllers, so the processor hardware and memory required for high-level MAC functionality need to be prepared only in the cell controller. There is. In practice, system-wide capabilities such as WEP coding and other special features can be increased at a reasonable cost by using a high-performance board-level personal computer or even a host computer as a cell controller. Can be done.

【0031】

By removing high-level MAC functionality from wireless access locations, the cost of equipment installed at those locations can be significantly reduced due to the lower processor power and capacity. In relation to related and mobile functions, RF port 18 prepares a beacon signal in response to a command generated by cell controller 14. When the mobile unit initiates the associated sequence, RF port 18 relays the associated message between the mobile unit 20 and the cell controller 14 during the associated process handled by the cell controller 14.

【0032】

In connection with the message traffic from the network processor to the mobile unit 20, the message packet is routed by the relay hub 12 to the cell controller 14 responsible for the mobile unit 20 with the destination. The message is stored in a buffer by the cell controller 14 and formatted, and in the preferred apparatus, the cell controller 14 encapsulates it as a mobile unit packet in a wired network packet destined for the responsible RF port 18. This packet is routed to RF port 18. The RF port 18 extracts the mobile unit packet from the message and transmits the packet as a radio signal to the mobile unit 20. RF port 14 also performs CRC calculations and generates CRC data to accompany the message. The mobile unit 20 responds to the RF port 18 with an approval response signal, and the RF port 18 generates an approval response status message and sends it to the cell controller 14.

【0033】

In connection with the message for the system connected to the communication line 16, the mobile unit 20 transmits the packet as a radio signal to the RF port 18. RF port 18 filters received message packets according to the BSS (basic service set) identifier in the packet, and if the packet has a BSS identifier associated with RF port 18, it performs a CRC check when the packet is received. .. The RF port 14 then generates an approval response signal and sends it to the mobile unit 20 and sends the received packet to the cell control device 14. The cell control device 14 stores the packet in a buffer, parses it, decodes it as necessary, and routes it to the host on the communication line 16 through the relay hub 12.

【0034】

The device at RF port 18 may be identical to the current access point used in the Spectrum 24 system, with some of the access point's software failing. RF ports are preferably simplified to reduce cost and power consumption. To reduce equipment costs, RF port power is supplied via an Ethernet cable, which also connects RF port 18 to a relay hub 12 or cell controller 14. The RF port 18 can be installed in a small container (eg, portable radio size) with an integrated diversity antenna, or with adhesive tape or Velcro for easy installation. An Ethernet cable is used for connection to the relay hub 12, and DC power is prepared by using a choke circuit such as the pulse model PO421 described in the international patent application referred to by the applicant. The choke circuit may be built into the Ethernet connector and is available in that form.

【0035】

RF port 18 requires low levels of processor power, software support, memory, and power consumption because it does not need to filter Ethernet addresses and perform 802.11 related and mobile functions. In the embodiment shown in FIG. 3, the RF port 18 only includes a digital signal processor (DSP) 28 that includes an internal random access memory (RAM) and ROM. The DSP38 may be one of the TMS320 series of DSP processors manufactured by Texas Instruments, such as the 5000 series, in particular the TMS320UC5402 or TMS320VC5402. This DSP prepares the interface between the Ethernet cable 46 and the RF module 42 in RF port 18, as shown in FIG. The RF module 42 is provided in housing 36 with one or more antennas 44 with DSP 38 and DC / DC power supply 40. The RF module 42 includes a 3860 or 3861 baseband processor such as the HFA3860B and serves as an interface for the digital portion of RF port 18, especially DSP38. In one configuration, the DSP38's ROM memory can be prepared with a "bootloader" firmware, which downloads the required DSP software instructions from the cell controller 14 at the same time as the RF port 18 starts and sends the instructions to the DSP38. Load into RAM.

【0036】

Currently preferred processors for low-level MAC engines are the TMS320UC5402 and TMS320VC5402. These parts are functionally identical but have different power consumption (VC5402 is currently in production, but UC5402 is still in the sample stage). The basic configuration of UC5402 / VC5402 is as follows. -100 MIPS execution speed -8 KB per chip ROM (composed of 4Kx16 bits) -32 KB per chip RAM (composed of 16Kx16 bits) -1 Two 16-bit timers with microsecond or better resolution -Two high-speed, fully-duplicated serial ports with smart DMA communication path support (each up to 50 megabits per second) -One fast 8-bit wide host / parallel port (160 megabits per second) -Six general-purpose DMA communication paths -16-bit external memory / I / O bus with internal wait state generation -Worst case of 3 directives (30 nanoseconds) 16 interrupts with latency -0.54 milliwatts / megahertz power consumption (30 milliamps at 1.8 volts at 100 megahertz) -Low power mode (6mA, 2mA, 2mA depending on settings) -Internal PLL that generates system clock with external crystal oscillator [0037]

This section describes how to use the 5402 DSP38 as a MAC engine for 802.11DS systems at 11 megabits per second. Obviously, it could be used in FH systems as well. The following focuses on how this 5402 acts as an interface to the Intersil 3860/1 baseband processor of RF Module 42 and how it performs low-level MAC functions.

【0038】

The first issue is how the 5402 DSP38 acts as an interface to the 3861 (most of the discussion applies to the 3860 as well) and the rest of the RF module 42. As shown in FIG. 4, the 3861 processor 53 in the RF module 52 of RF port 50 has two main interfaces, which are in series. The first interface, named DATA, is used to transfer data between the MAC engine, including DSP64, and the 3861. It has four lines TxD, TxC, RxD, RxC and operates at a maximum speed of 11 megabits per second. The exact speed depends on the packet transmission speed. The clock signals for both interfaces are generated by the 3861 and therefore the transmission is controlled by the 3861. Both can be stopped at any time with the 861 and the speed can be changed. A second serial interface, named CONTROL, is used to load commands into the 3861 and read status information from the 3861. This interface is a 4-wire 2-way interface that uses 1 data line, 1 clock line, 1 "direction control" line, and 1 chip selection line. This series interface can also operate at speeds up to 11 megabits per second. The series interface also has additional control and status lines such as reset, TX_PE, RX_PE, TX_RDY.

【0039】

The 5402DSP38 has two sets of fully dual series interfaces that can operate up to 50 megabits per second (as a 100 MHz clock). They can be ticked using internal or external supply devices. In this design, one of a set of serial interfaces named SER1 can be used to connect to the high speed data line of 386 interface 53. The 5402DPS38 interface has the same baselines (RxD, RxC, TxD, TxC) as the 3861, so they are connected with minimal effort. The 5402 uses 1.8 volts for its core, but its input and output lines have a tolerance of 3.3 volts, so it can function as an interface to the 3861 without a converter. Moreover, since they are completely static, clock line start / stop operations can be handled from the 3861.

【0040】

Data transfer will take place under DMA control within 5402 using what Texas Instruments (TI) calls "automatic buffer mode." It has essentially a dedicated DMA communication path for each serial port interface (2 DMA communication paths per serial port interface). Since these communication paths access the banks of SRAM that operate independently of access, the transfer does not impact the operation of the CPU. The CPU can initiate transfers in either direction and will be notified by an interrupt when they are complete.

【0041】

Interface functions for the control serial port on 3861 interface 53 can be performed in three different ways. The first method, shown in Figure 4, uses a second serial port named SER2 on a 5402 DSP64 with a small amount of combinatorial logic / buffer, with a single data line of 3861 and a dual data line of 5402. Perform a conversion between. Another method is to use an external shift register for serial / parallel conversion. This register is on the I / O bus of 5402 and is loaded / read by 5402 and the data is shifted between it and 3861. The third method is to use an external buffer / latch on the 5402 I / O bus to "bit bang" the clock / data line to 3861. The second and third methods open the second series communication path for yet other applications such as providing a high speed series interface such as Ethernet or USB, which is preferred over the first method in some applications. May. The cost of all solutions is about the same, as they all require a small amount of external combinatory logic. The same logic could be applied to interface with synthesizers. Synthesizers aren't accessed as often as the 3861's control port, so a "bit-bang" method would work.

【0042】

Finally, interfacing to the various control and status lines represented by 3861 will be done through simple two-way registers / latches connected to the 5402 I / O bus. The 5402 can read / write this register because it needs to control and monitor the 3861. All control / monitoring functions (including the serial control interface) could be combined in one 16-bit buffered register latch. A parallel control / status line will be connected to a particular line in this latch. A series control interface will also be connected and "bit banged" when needed to move data between 5402 and 3861.

【0043】

The device shown in Figure 4 uses a crystal CS8900A controller 63 coupled to the DSP64's parallel port to serve as an interface to Ethernet port 58. The Ethernet connector / choke 58 accepts the cable 60 and supplies DC power from the cable 60 to the DC / DC power supply 62. RF port 50 in FIG. 4 includes spaced diversity antennas 54 and 56 to improve reception in multiple path conditions.

【0044】

The premise of this design is that TI's DSP can implement all low-level MAC functionality without the assistance of external hardware. Of course, this is the most demanding model, but you'll find that the 5402 can withstand this task. The jobs that require the most computing power are CRC32 and WEP processing. The CRC32 calculation is performed over all packets, on time to generate an acknowledgment (ACK) if the CRC is found to be correct (or to attach the calculation result to the packet leaving the transfer). Must be completed. This means that the CRC calculation needs to be done in near real time during the packet transfer between 3861 and 5402. TI has shown in the application memo that CRC calculations can be performed with 13 directives by the 5000 series DSP. At 100 MIPS, this is about 130 nanoseconds. At 11 megabits per second, it takes about 70 nanoseconds to transfer one byte, so there is plenty of time to do the CRC. When receiving the packet, the serial port will forward the data from 3861 to the SRAM in 5402. At the same time, the CPU in 5402 will read the received bytes from SRAM and calculate the CRC. Of course, you'll need to make sure it didn't overrun the receive buffer, but that would be a relatively easy task. Almost the same processing will occur during the transfer. In either case, the CPU has enough time to do the CRC.

【0045】

WEP processing is a more difficult function than CRC32 because it includes both RC4 coding and CRC32 when performed within RF port 50. At the same time, it does not need to be completed prior to ACK generation / reception and is not done for all packets (data packets only). The RC4 coding function consists of two parts. Generating a code table (256-byte table) using the select key and executing the coding / decoding process. Based on the sample code, table generation would require about 1200 directives (12 ms at 100 MIPS) and the coding / decoding process would require about 12 directives per byte. .. This cost remains the same for 40 or 128 bit keys. WEPCRC32 would require a separate 13 directives per byte.

【0046】

The computational load per byte of WEP would therefore be about 250 nanoseconds with 25 directives or 100 MIPS. When added to packet CRC32, the total load is 38 directives per byte. As pointed out, about 77 commands per byte are available at 11 megabits per second, so about 50% of the CPU is spent on CRC / WEP tasks. The biggest issue is the 1200 clocks (12 microseconds) required to generate the code table during reception (for forwarding, the calculation is completed before packet forwarding begins). .. Pausing to generate a table would delay the CPU by about 18 bytes (12 microseconds at 770 nanoseconds per byte) in CRC / WEP / CRC processing. Therefore, 40 data bytes (1200 clocks / extra 30 clocks per byte) will be required to catch up with both packet CRC and WEP / CRC functions. The minimum TCP / IP header is at least 40 bytes (plus any user data), so there should be plenty of time. In any case, there is no harm in the WEP / CRC calculation with a slight delay. An alternative method is to catch up with the CRC calculation first and then with WEP / CRC.

【0047】

After CRC and WEP / CRC processing, the second most important activity is parsing the header on reception and generating it on transfer. This is because the base station needs to identify the packet and generate an appropriate response. Upon receipt, the processor must parse two or three 48-bit addresses and at least one 16-bit header command field. It may be necessary to generate an ACK when the packet is complete. The 5402 can easily handle these functions. Since these functions are executed before WEP processing, the CPU has 64 directives (77-13) per byte to execute these functions. Since many of them can be executed based on 16-bit or even 32-bit bases (5402 supports both 16-bit and 32-bit operations), there can be 128 or 256 instructions per data item (there are 128 or 256 instructions per data item). That is, 256 commands to perform 32-bit address checking). These functions are performed in 2 megabits using a 1MIPS188 CPU. Applicants have a 100 MIPS CPU that performs the same task at 11 megabits per second.

【0048】

ACK generation is also relatively simple. The ACK frame is only 14 bytes long and contains 4 CRC-32. Given a long (80 microseconds) preamble, there would be 8000 directives to prepare the ACK. The same applies to RTS / CTS exchanges. There are two 16-bit timers available on the 5402. In this model, one will be used for TSF timing and the other will be used for all other features. In practice, there are only a few other timer functions, such as NAV, reforwarding, and collision avoidance slot countdown. Retransfer and collision avoidance activities are continued only to initiate retransfer while waiting for an ACK or after detecting an unused network. In such cases, there is no ongoing data transfer and therefore there are plenty of CPU cycles available.

【0049】

Support for the MUPSP feature comes in a variety of ways, depending on how much external hardware, if any, is prepared. The 5402 provides various means of saving power. The first is simply to slow down the CPU clock through a software controlled PLL within the unit. The 5402 generates an internal clock through an external crystal oscillator or a clock-driven PLL. The PLL multiplies the fundamental frequency of the crystal oscillator / external clock by a coefficient determined by the software. So one way to control power consumption is to simply slow down the CPU clock. Since the CPU portion of the processor consumes most of the power, slowdowns have the greatest impact on power consumption.

【0050】

The second method is to use one of the processor's IDLE modes. IDLE1 completely stops the CPU clock, but keeps everything else running. Power consumption in this mode is around 6mA at 100MHz. The CPU can be restarted by any interrupt (internal or external). At IDLE2, the system clock is stopped and consumption is reduced to 2mA. At IDLE3, all system functions are stopped and consumption is reduced to about 2 microamps. In all cases, all conditions are maintained. An external interrupt is required to restart the CPU with IDLE2 and IDLE3. In such cases, an external low power timer may be needed. Therefore, without external hardware, power consumption could be reduced to at least 6mA, and perhaps less. With a simple external timer, it could be reduced to microamps.

【0051】

In short, the huge CPU output of the 5402 allows all low-level MAC functions to be performed by software. Moreover, it has sufficient output and memory to handle additional "high level" functions such as packet transfer, splitting, and reassembly that can also be performed on the cell controller. The system 10 of the present invention conforms to IEEE standard 802.11 and will therefore work with any mobile unit 20 including existing units conforming to that standard. However, the improvements made to RF port 18 to reduce the complexity and cost of these units are mobile units with sufficient main processor capacity to handle mobile unit functions that correspond to higher levels of MAC functionality. It can also be applied to 20.

【0052】

Referring to FIG. 2, a block diagram of a mobile unit computer and a mobile unit 20 with a WLAN adapter 24 connected to it to prepare wireless communication to system 10 of FIG. 1 is shown. In the mobile unit 20 of FIG. 2, the low level MAC function is performed within the WLAN adapter 24, which also includes the RF module 28 and the antenna 29. The form of the WLAN adapter 24 may be similar to that of existing adapters, but preferably the adapter 24 is simplified to perform only the low-level MAC functions of the IEEE standard 802.11 protocol and is special within the host computer 22. Software 34 has been simplified to perform high-level MAC functions such as association and movement. In a preferred configuration, the MAC function of the adapter 24 is performed on the digital signal processor 26 as described below, which may be the same type of DSP as described in relation to RF port 50.

【0053】

This section describes how the 5402 DSP can be used as a MAC engine in a mobile unit configuration. Two things are considered when building a MU WLAN solution. The first is the location of those MAC functions, and the second is the physical interface to the host. The position of the upper level MAC function can vary considerably. The possibilities are: -Full functionality on MAC engine DSP processor 26 -Full functionality on host processor 22 -Moving / related features on host processor 22 and remaining features on MAC engine 26 -Move / related / retransfer functionality on host 22, remaining functionality on MAC engine 26 [0054]

High-level MAC feature location selection is MU It has a significant impact on the cost of WLAN adapters. If you want to put at least some high-level features on the host processor 22, you could just outperform the 5402 on the WLAN adapter. Functions that can be placed on the host would be mobility and related controls. High-level features such as transfer and split / reassembly could be left on the 5402. This split allows considerable savings as it does not require a separate processor / memory subsystem on the WLAN adapter. There are two reasons why not all MAC features are on the 5402. First, the 5402 memory space only has 32 kilobytes of SRAM for both code and data. In the execution of some MACs, such as frequency hops, the code space alone exceeds 32 kilobytes. The second reason is that the software on the 5402 is for satisfying difficult and real-time tasks such as CRC and WEP processing. Attempting to add software-intensive tasks would simply complicate the process. If we needed another processor, such as ARM or perhaps a second 5000 series processor, we could have added top-level functionality to it.

【0055】

Instead, all MAC features could be put on a faster and / or larger version of the 5402 processor. Such processors will probably have higher clock speeds (current 5000 series components can time up to 160 MIPS) and more memory (eg 64 KB instead of 32 KB). Let's go. Both the second processor and the faster / larger 5402 will incur extra costs and consume extra power.

【0056】

This section describes one way you can place a MU WLAN adapter for various hardware host interfaces using the 5402. Assuming a sufficient amount of high-level MAC functionality has been transferred to the host processor, the PLAN adapter only requires 5402. A second processor could be added for any of the solutions outlined below. In all of the following solutions, it is assumed that the routine code for 5402 is loaded from an external supply device (such as computer 22) through host interface 32. This eliminates the need for flash memory on the adapter board and saves a few dollars from the process. It should be pointed out that the 5402 is associated with an 8 kilobyte mask programmable ROM, which will contain the bootloader program (required for USB and Ethernet host interfaces). The bootloader will be smart enough to download run-time code directives on any serial interface available.

【0057】

The simplest of all interfaces would be the host port on the 5402 used by the host. This port acts as a dual port interface to memory in 5402. It may not be a standard interface, but it's quite suitable for dedicated systems. With it, computer 22 can read and write to memory randomly or in sequence. It is an 8-bit interface and can operate at 160 megabits per second. When operating in random access mode, computer 22 uses two writes to the port to generate a 16-bit address and then performs a read or write operation. Such a mode allows the host to set command blocks and the like in the memory of the 5402. Sequential mode allows the host to move data in and out of 5402 memory very quickly (approximately 160 megabits per second). This will be used for data transfer. If this method were used, the only digital component of the WLAN adapter would be the 5402.

【0058】

In the system of FIG. 1, the cell controller 14 is preferably a board-level personal computer coupled to the relay hub 12 by 10-megabit and 100-megabit Ethernet ports. For smaller systems, a 350 MHz Pentium computer with 16 MB RAM may be used. For large systems with many RF ports, a 500 MHz Pentium with 64 MB RAM is suitable. Communication with the wired network is preferably performed at 100 MHz. Communication with the RF port can be performed at 10 MHz. A second cell controller may be supplied for large systems and / or for backup in the event of one cell controller failure. Reliability can be improved by providing a dual fan and a dual power supply. Flash disk memory may be used for reliability. Alternatively, the cell controller 14 may be built into the relay hub 12 or the host processor.

【0059】

The operating system for the cell controller 14 may be a real-time operating system such as VRTX or QNX, which provides multiple task processing, a complete network stack, and ancillary equipment. A Java-based, web-based management ancillary equipment on the customer side will be prepared to maintain the status form of the cell controller 14, RF port 18, and mobile unit 20. The cell control device 14 includes an application that provides mobile unit related management, movement, and packet buffer management. These applications are similar to those running on the current access points of the Spectrum Spectrum 24 system. The cell controller 14 may also provide QoS support, user approval, and configuration management. By placing these features on a personal computer, the cell controller facilitates system management and program updates using available programming tools. Further, approval or modification of the management function is only required to be placed in the cell control device 14, and modification of the software of RF port 18 is not required.

【0060】

The cell control device 14 handles routing of all sent and received messages to and from the mobile unit. The cell control device stores the message packet received from the wired network in a buffer, determines the appropriate RF port 18 to which the destination mobile unit 20 is associated, and forwards the packet to the RF port 18. The cell controller 14 further performs WEP coding / decoding and associated CAC. The cell controller 14 may also have additional functions such as maintaining firmware and downloading to RF port 18. After powering up, RF port 18 sends a download request to cell controller 14 using the bootloader routine stored in ROM. The cell controller then downloads the firmware, including channel allocation and morphological information such as ESS and BSS identification, to RF port 18. The cell controller 14 and the RF port 18 further share a common TSF clock.

【0061】

The mobile unit computer 22 of the mobile unit 20 comprises software that performs high-level MAC functions similar to those outlined above. Advantageously, the software 34 can be programmed using the same operating system that the computer has, which allows the user to have a familiar user interface such as Windows®. it can. Mobile unit software 34 prepares header construction, movement, and related MAC functions. The mobile unit computer 22 may also download the firmware to the processor of the WLAN adapter 24.

【0062】

As is clear from the above description, the hardware for RF port 18 and WLAN adapter 24 of mobile unit 20 can be substantially similar except for the possibility of host interface to Ethernet or mobile unit. Is. Further, the logical cell controller function and the higher-order MAC function executed by the mobile unit host processor can be executed on any computer system. Using the RF port 18 of the present invention coupled to a computer system, it is possible to provide either a mobile unit or a wireless network according to the software provided. Software for RF port 18 can be downloaded from the host system, so a simple combination of a computer and one or more RF ports is a WLAN mobile unit or WLAN by supplying feature-selectable firmware to the RF port processor. It can function as a host, or both.

【0063】

In the device of FIG. 5, the personal computer 70 is equipped with software 72 and is connected to one or more RF ports 50A and 50B to prepare a complete host system for wireless data communication. With this configuration, for example, office equipment is connected to server 70 via a wired network for traditional LAN operation, and one or more RF ports 50 are also connected to server 70 in the LAN system in a small enterprise. It could result in data communication between the 70 and the mobile unit. The server can perform higher MAC functions and download firmware instructions to the RF port. Alternatively, the firmware directive can be placed on the PROM of the RF port.

【0064】

FIG. 6 shows a device that wirelessly accesses the Internet using the RF port 50 of the present invention. Internet access to modem 82 on line 80 may be provided by cable, data set label (DSL), or fiber optic transmission. The RF port 50 may be prepared to include MAC firmware on the PROM or may be configured to include a boot loader program to download the firmware from the ISP server. When installed in a home or office, the mobile unit 20 can initiate Internet access in association with RF port 50. The ISP server may perform high-level MAC functions, or those functions may be provided on RF port 50. The mobile unit 20 may be a home or office personal computer 22 equipped with a WLAN adapter 24 as shown in FIG.

【0065】

FIG. 7 shows examples of communication formats that can be used in various embodiments of the present invention. The example of FIG. 7 assumes that this configuration includes a host 90 connected to a dedicated cell controller 14, which is also connected to RF port 18. It should be clearly understood that the functions of the logical cell controller may be performed on host 90, especially in simple systems. In the example of FIG. 7, the host 90 sends a message A having 100 data bytes to the cell controller 14 through the Ethernet packet 100. Packet 100 has a destination address of a mobile unit (M1) and a source address of a host (H), and includes data (A). The cell controller 14 formats the data in 802.11 format, including the destination corresponding to the mobile unit (MU1) 20. The cell encapsulates this 802.11 packet containing data A in an Ethernet packet 104 destined for RF port 1 (RF1) from the cell controller (CC).

【0066】

The RF port 18 receives the Ethernet packet 104 from the cell controller 14, generates the 802.11 format RF packet 112, and sends it to the mobile unit 20 together with the data A. 802.11 header generation can be prepared on either cell controller 14 or RF port 18, but packet 104 contains mobile unit identification data as an 802.11 header, or otherwise RF port 18 It should be understood that the mobile unit identification data must be included in order to be able to generate the header. RF port 18 also performs a CRC calculation and adds the result to 802.11 packet 112.

【0067】

The second message "B" with 1500 bytes of data is also shown to be derived as an Ethernet packet 102 destined for cell controller 14 from host 90. The cell controller divides the data message B into three fragments B1, B2, and B3 to fit the data limit of 500 bytes for each 802.11 packet. These fragments are sent to RF port 18 as Ethernet packets 106, 108, and 110, which sends RF signal packets 114, 116, and 118 to mobile unit 20.

【0068】

The same applies to reverse communication. Message C has 100 bytes and is transmitted from the mobile unit 20 to the RF port 18 as an 802.11RF signal packet 200. The RF port 18 encloses this message in the Ethernet packet 208 and transmits it to the cell control device 14, and the cell control device 14 extracts the destination information and data and supplies the Ethernet message 216 to the host 90. The larger message D is sent to RF port 18 as message fragments 202, 204, and 206, relayed to cell controller 14 as Ethernet packets 210, 212, and 214, and hosted 90 as reassembled Ethernet packets 218. Will be sent to.

【0069】

Although claims have been described for the applications identified above, one of ordinary skill in the art understands that other and further modifications may be made without departing from the scope of the invention. It is intended that all such changes and modifications will be claimed to be within the true claims of the invention.

[Simple explanation of drawings]

[Figure 1]

It is a block diagram of the wireless communication system by this invention.

[Figure 2]

FIG. 6 is a block diagram showing an example of a mobile unit arranged for use in the system of FIG.

[Fig. 3]

FIG. 6 is a block diagram showing an example of an RF port arranged for use in the system of Figure 1.

[Fig. 4]

It is a more detailed block diagram of the preferred embodiment of the RF port according to the present invention.

[Fig. 5]

It is a block diagram which shows the arrangement of the computer and RF port which brings about the simplified wireless local area network by this invention.

[Fig. 6]

It is a block diagram which shows the arrangement which prepares the wireless access to the Internet using the RF port of this invention.

[Fig. 7]

It is a figure which shows the signal format by one Embodiment of this invention.

[Explanation of symbols]

10 Wireless data communication system 12 Relay hub 14a cell controller 16 Wired network 18a Simplified RF port 20a mobile unit