Method and apparatus for RFID communication

Abstract

A plurality of battery-operated transceivers encapsulated by lamination to form a sheet of independent transceivers is tested in a two piece fixture that forms an enclosure surrounding each in-sheet transceiver. Each enclosure has an antenna for transmitting a command signal to the transceiver at a known power level and for receiving a reply message from the transceiver containing a power level measurement made by the transceiver. Test methods using the fixture of the present invention are also described. An rfid tag and interrogator may each include a transmitter and a receiver. The tag and interrogator may communicate with each other at different frequency bands and may communicate in accordance with a wireless communication protocol.

Description

An RFID tag and interrogator may each include a transmitter and a receiver. The tag and interrogator may communicate with each other at different frequency bands and may communicate in accordance with a wireless communication protocol. The following disclosure corresponds to the Detailed Description Section and figures of U.S. Pat. No. 5,365,551 (the '551 patent) incorporated by reference above. FIG. 1 of the '551 patent is a functional block diagram of communication system 30 of the present invention as described in the '551 patent. In FIG. 1 of the '551 patent, commander stations 10 and 34 and responder stations 40 and 36 are coupled to common medium 32 by network interfaces 26 and 60, respectively. In practice, a plurality of commander and responder stations would be distributed geographically. The type of medium selected for communication depends on the communication system application; see below for equivalent variations. The embodiment depicted in FIG. 1 of the '551 patent illustrates the invention in an application such as airport baggage handling. For this embodiment, the medium is free space through which radio frequency communication are transmissable.

FIG. 8 of the '551 patent is a table that describes several responses and refers to response formats described in FIG. 9 of the '551 patent. As shown in FIG. 9 of the '551 patent, response formats 192-196 include LOCAL ID. ARBITRATION NUMBER, and TAG, which have the meanings already described above. By including LOCAL ID and ARBITRATION NUMBER in each response. in cooperation with locked bit 88 one responder station can respond unambiguously to one commander station in the presence of a plurality of commander and responder stations. The INVERTED ARBITRATION NUMBER in format 192 is the binary ones-complement of the ARBITRATION NUMBER and is included for increased accuracy of communication. REVISION in format 192 is a one-byte value set by a communication system developer at the time of manufacture or commissioning of a responder station. REVISION represents the responder station configuration and connotes its capability. STATUS in format 196 is a one-byte code chosen by responder station 40 to convey current conditions of important system events such as low battery. uncorrectable data received. write protection. And similar information which may indicate to commander station 10 that communication should be repeated or abandoned. DATA in response format 194 includes some or all of the contents of any or all devices including memory 64, register array 66, flag register 84, or random number generator 90. 
A communication system, according to the present invention as described in the '551 patent, includes commander and responder stations that adhere to a method of communicating called a protocol. In general, the protocol of the present invention as described in the '551 patent places different requirements on a commander station than on a responder station. Thus, there is a commander station method (FIG. 10 of the '551 patent) and a responder station method (FIG. 11 of the '551 patent). These methods together implement the communication system protocol. 
Operation according to the present invention as described in the '551 patent produces the following characteristic effects at the system level. First. a commander station will not begin transmitting during the transmission by another commander station or by a responder station. Operation, according to the present invention as described in the '551 patent, does not prevent more than one commander station from beginning transmission simultaneously; however, it is feasible to couple commander stations to a second medium or to constrain commander stations to a second or expanded protocol on common medium 32. For example, commander stations 10 and 34 include personal computer system 12, which can be augmented with a peripheral controller for operation over ethernet. Communication over the second medium can be used to prevent simultaneous broadcast over common medium 32. For example, a second protocol on common medium 32 may include operator action to assign time slots, back off delays, or similar means for media access whether central or distributed. Several embodiments for these means for media access have been described by Stallings in his work already incorporated by reference above. 
Second, a responder station will not transmit unless it has first received a command to which it determines it must respond. The response is made within a predetermined time immediately following receipt of the command. 
Third, a commander station can form a command in a manner calling for all, more than one, or one responder station to respond. An important object of the communication system protocol in a communication system of the present invention as described in the '551 patent, i.e. uninterrupted communication, is achieved after a commander station determines how to cause only one responder station to respond. The program flow diagram of FIG. 10 of the '551 patent and the state diagram of FIG. 11 of the '551 patent describe how uninterrupted communication between one commander station and each responder station is achieved when a plurality of commander stations and a plurality of responder stations are simultaneously coupled to a common medium. 
FIG. 10 of the '551 patent is a program flow diagram of the protocol followed by a commander station of the present invention as described in the '551 patent. A practical example of a communication system will be used to describe the flow diagram. 
In a communication system for airport baggage handling the quantity and identity of responder stations within the radio communication range of a commander station varies over time. A commander station may be at a fixed operator station within radio range of a moving belt conveying baggage toward a V-junction of conveyors. When baggage tags are constructed as responder stations and when each tag has destination information stored in memory 64, the commander station, through communication with each baggage tag, can control the routing of each bag through the junction onto one of two conveyors. Assume that each responder station also has information in memory 64 describing its sequential position on the conveyor. Such a sequence number could be a date and time of day when the bag passed through a chute upstream of the commander station. 
As a group of bags approaches the commander station, the commander station has a fixed amount of time to determine the identity of each responder station, in order to establish uninterrupted communication. For proper baggage handling, the commander station must routinely and repeatedly identify all bags on the conveyor. To do so, at FIG. 10 of the '551 patent block 210, commander station 10 specifies a group of responder station addresses by choosing values for BRANCH and MASK. BRANCH and MASK values are determined in a manner to be explained by reference to FIG. 12 of the '551 patent below. In one embodiment, the initial group specification, i.e. BRANCH and MASK values, would specify all possible responder stations. Commander station 10 at block 212 generates an “identify, clear, and generate” (IDCG) command having a format according to FIGS. 4,5, and 6 of the '551 patent. When the media is clear to broadcast, block 214, as indicated by OK-to-transmit signal 116, the IDCG message is broadcast, block 216. An IDCG message causes each responder station that is a member of the group to clear locked-bit 88, generate a random number and retain it as its ARBITRATION NUMBER, and broadcast a response. The responder station's reactions to 10, lOG, IDC, and IDCG commands are explained further in reference to FIG. 11 of the '551 patent below. 
Commander station 10 now loops through blocks 220 and 218 for a response to be received as indicated by OK-to-transmit signal 116 or a time out elapsed condition. If a response was received, as indicated by a false state of OK-to-transmit signal 116, commander station 10 at block 222 determines whether a collision occurred. as indicated by a false state of proper-message-received signal 132. If commander station 10 determines that a collision occurred. it will determine at block 224 whether all possible members of the initial group of responder station addresses specified at block 210 have been addressed in an lD, lDG, IDC, or IDCG command. How this determination is made will be further explained with reference to FIG. 12 of the '551 patent below. If all subgroups have not been tried, the commander station again specifies a group of responder station addresses, for example, a subgroup or disjoint group of a prior group. At block 228 commander station 10 generates an 10 command according to FIGS. 4, 5, and 6 of the '551 patent and continues the method from block 214. 
If, at block 218, a predetermined time elapsed without a false condition appearing on OK-to-transmit signal 116, commander station 10 concludes that no response was transmitted and continues the method at block 224. 
If, at block 222, the proper-message-received signal is true, then commander station 10 concludes that only one responder station responded. At block 230, commander station 10 determines and validates the responding responder station's ARBITRATION NUMBER according to response format 192 using ARBITRATION NUMBER and INVERTED ARBITRATION NUMBER. According to a particular system communication objective, commander station 10 now selects a command from FIG. 5 of the '551 patent which will cause the responder station to set its locked-bit 88. For determining baggage destination and positional sequence on the conveyor, commander station 10 could select RD. Using the appropriate command format shown in FIG. 6 of the '551 patent, commander station 10 generates a message at block 232, loops until the OK-to-transmit signal indicates that the medium is clear to broadcast at block 234, then broadcasts the command at block 236. Commander station 10 again awaits a proper response message by looping at block 238 through block 240. If a predetermined time elapses at block 240, commander station 10 continues the method at block 234. If a response is received without error at block 244, as indicated by proper-message-received signal 132, then two party uninterrupted communication between commander station 10 and one responder station 60 has been established. Further communication may be required, as indicated by the STATUS code in the received response format 192 or to accomplish other system communication objectives. 
It is possible at block 224 for the commander station to determine that no further subgroup, super group, or disjoint group remains to be commanded using the 10 command. Suppose, for example, that all practical values of BRANCH and MASK have been used. If the immediately preceding broadcast at block 216 elicited no response at block 218, then commander station 10 can conclude that all responder stations have been identified. Otherwise, at block 248, commander station 10 generates an identify and generate command (lOG) according to the format in FIGS. 4, 5, and 6 of the '551 patent using the same group that was specified in block 210. Commander station 10 continues the method at block 214. 
Although the same group is specified, a responder station that has been identified at block 244 will not respond, since its locked-bit 88 has been set. Collisions are less likely to occur with each pass through the loop from block 214 to block 248 because a smaller number of responder stations can respond. Hence, the method of FIG. 10 of the '551 patent converges toward identifying all responder stations. The communication system designer must select the precision of BRANCH and MASK values to assure conversion within system time allowances, for example, 8-bit BRANCH and MASK values are compatible with conveyor speeds and radio ranges needed for airport baggage handling. 
FIG. 11 of the '551 patent is a state diagram of the protocol followed by a responder station of the present invention as described in the '551 patent. Responder station 40, begins in idle state 310 when power is applied or restored according to wake-up signal 174. In part, the idle state is indicated by contents of command register 56 not corresponding to a valid command. The idle state is re-entered to interrupt command processing when improper-byte-received signal 180 is raised by receiver logic 178. A valid command loaded into command register 56 causes state transition to address check state 312. 
In address check state 312, microsequencer 42 determines whether responder station 40 has been addressed by one of two methods. First. if the command conforms to format 142, the responder station is addressed when the result of ARBITRATION NUMBER logically ANDed with MASK is bitwise identical to BRANCH. ARBITRATION NUMBER is the current contents of a particular register in register array 66. MASK and BRANCH are values received in the command and stored in register array 66. Logical operations and comparisons are performed by ALU 72 which produces A=B signal 82. If A=B signal 82 is not asserted, state 314 is entered. Responder station 40 may remain in state 314 until a predetermined time elapses. Responder station 10 re-enters idle state 310, after the predetermined time elapses. 
To illustrate the importance of such a delay, suppose that commander and responder stations employed radio transceivers for network interfaces 60 and 26. Then, suppose responder station 40 is within range of two commander stations 10 and 34, but commander stations 10 and 34 are out of range of each other. When commander stations 10 and 34 validly produce back to back commands, the delay interposed by state 314 prevents responder station 40 (not addressed by commander station 10 in the first occurring command) from responding to commander station 34 in the second occurring command. Without the delay, a collision could occur that may confuse commander station 10. 
A second way to determine whether responder station 40 has been addressed applies for commands having formats 194 and 196. Accordingly, responder station 40 is addressed when ARBITRATION NUMBER, retained in register array 66, is bit-wise identical to ARBITRATION NUMBER as received in the command. Comparison is performed by ALU 72 which produces A=B signal 82. If A=B signal 82 is not asserted, state 314 is entered as already described. Otherwise, transition is made to decode state 316. 
Decode state 316 follows address check state 312 in response to A=B signal 82. If the command opcode is not recognized then no response state 314 is entered. For some commands, a further condition such as the state of locked-bit 88, if not met. will cause the command to be treated as not recognized. Each recognized command opcode causes microcode execution to begin at a section of microcode for the purpose of directing microseqencer operations to process the particular received command. Four commands are illustrated as separate states 318 through 324 and other commands are illustrated in summary by pseudo state 326. 
When the opcode for command IDCG has been received, state 318 is entered for identify, clear, and generate operations. An lOR response (according to FIGS. 8 and 9 of the '551 patent) is selected, locked-bit 88 is cleared, the content of random number generator 90 is stored in register array 66 as ARBITRATION NUMBER, and transition is made to state 328. 
When the opcode for command lDG has been received and locked-bit 88 is not set, state 320 is entered for identify, and generate operations. An lOR response is selected and a new ARBITRATION NUMBER is generated as already described for state 318. Transition is then made to state 328. 
When the opcode for command IDC has been received, state 322 is entered for identify and clear operations. An lOR response is selected and locked-bit 88 is cleared. Transition is then made to state 328. 
When the opcode for command 10 is received and locked-bit 88 is not set, state 324 is entered for an identify operation. An lOR response is selected. Transition is then made to state 328. 
When the opcode for other commands (including RD and WD) is received, lockedbit 88 may be set and other functions may be performed. Other functions may include writing data to memory 64, writing data to register array 66, altering the state of registers including flag register 84, and other operations controlling responder station configuration and operation. Transition is then made to state 328. 
Upon transition to state 328, the response selected by a prior state is generated according to FIGS. 7, 8, and 9 of the '551 patent and broadcast. In one embodiment, the response is broadcast as it is being generated. Transition to idle state 310 is made, after broadcasting the response. Note that responder station 40 does not wait for clear medium prior to broadcasting the response. According to the present invention as described in the '551 patent, collision detection by responder stations is not necessary to accomplish uninterrupted communication. 
The ARBITRATION NUMBER generated by responder station 40 and the BRANCH and MASK numbers chosen by commander station 10 operate to establish uninterrupted communication. We now turn to a further explanation of the method used by commander station 10 to choose BRANCH and MASK values. 
FIG. 12 of the '551 patent is a binary tree diagram of BRANCH values and MASK values chosen by a commander station. A tree is a type of graphic representation. There are several types of trees known in mathematics and computer science. The tree depicted is a binary tree where a node can have two branches, shown descending left and right from a node. Each node corresponds to a unique combination of values for BRANCH and MASK, which are nbit binary numbers having the same precision. As illustrated, BRANCH and MASK are n-bit binary numbers. In a communication system for airport baggage handling, a-bit numbers would be used. The precision employed for BRANCH and MASK must be identical to the precision selected for ARBITRATION NUMBER generated by responder station 40. 
Recall that responder station 40 uses the expression ARBITRATION NUMBER AND MASK=BRANCH to determine if it is addressed, where ARBITRATION NUMBER is the value retained in register array 66 from random number generator 90. When MASK is 0 and BRANCH is 0 all values of ARBITRATION NUMBER satisfy the expression, i.e. all responder stations coupled to common medium 32 conclude they are addressed. On the other hand, if MASK has a ‘1’ bit in every position, then the expression is satisfied for only one value of ARBITRATION NUMBER. 
When MASK is arranged with ‘0’ and ‘1’ bits, the expression is satisfied by a group of values for ARBITRATION NUMBER, and potentially a group of responder stations could conclude they are addressed. Note for a responder station to be addressed, BRANCH at bit position ‘q’ must be ‘0’ when MASK at bit position ‘q’ is ‘0’, for all values of ‘q’. When MASK at bit position ‘q’ is ‘1’, BRANCH can take two values for that bit position which correspond to the left and right branches of a binary tree. 
At the first level of the tree, nodes 702 and 703, MASK is ‘1’ in bit position ‘r’. The corresponding bit position of BRANCH is ‘0’ at node 702 and ‘1’ at node 703. At the second level of the tree, nodes 704 through 707, MASK is ‘1’ at bit positions ‘r’ and ‘s’. For example, the value for BRANCH at node 707 is the parent node BRANCH value (001 at node 703) modified by forcing a ‘1’ (for the right-hand branch) at bit position's', hence 011. In like manner, the BRANCH and MASK values for any node in the tree can be determined. In FIG. 12 of the '551 patent MASK bit positions have been given in an order right to left. Any other order of bit positions would be equivalent. Methods for choosing first and subsequent values for BRANCH and MASK can now be explained in terms of traversing from node to node on a binary tree. 
When commander station 10 broadcasts a request for identification (an ID, IDC, lDG, or IDCG command) one of three events can occur. BRANCH and MASK values given at a particular node that represents a first group of responder stations. First, commander station 10 could receive no response from which it could conclude that no responder station in the first group is currently coupled to the common medium 32. Second, a proper response could be received. From that event, commander station 10 could conclude that only one responder station in the first group is currently coupled to common medium 32. Third, from an improper response received, commander station 10 could conclude that a collision of more than one response occurred. An improper response could be caused by, for example, weak coupling, high noise levels, or weak received signals. However, these causes can be treated in the same way as a collision to simplify the commander station protocol without substantially degrading system performance for applications including airport baggage handling. Therefore, an improper response simply merits further search. 
An efficient search for the identity of each of several responding responder stations is equivalent to an efficient search for the leaves of a binary tree. A leaf is a node having no further branches. When use of the values for BRANCH and MASK at a node produces no collision, the node is a leaf. Tree search methods are easily implemented using known computer programming methods. 
Tree search methods are essentially of two types, breadth first and depth first. A particular communication system application may use one method or the other to optimize commander station computing time and memory space objectives. An explanation of these methods using the programming language PASCAL is given by E. Horowitz and S. Sahni in “Fundamentals of Data Structures in PASCAL” pp 203-265 and 326-332 published by Computer Science Press Inc., Rockville, Md. (1984), incorporated herein by reference. 
Suppose that two responder stations 40 and 36 and one commander station 10 are coupled to common medium 32. The binary tree in FIG. 12 of the '551 patent illustrates a sequence of BRANCH and MASK values used by commander station 10 to identify responder stations. Timing diagrams in FIGS. 13 and 14 of the '551 patent illustrate the same sequence showing decisions at commander station 10 decision blocks (according to the commander station method of FIG. 10 of the '551 patent) and responder station control signals (according to the responder station method of FIG. 11 of the '551 patent) as commander station 10 establishes uninterrupted communication with each responder station. 
Beginning at FIG. 10 of the '551 patent block 210, FIG. 12 of the '551 patent node lOt and FIG. 13 of the '551 patent time 810, commander station 10 chooses BRANCH=OOO and MASK=OOO, calling for all responder stations to respond. At time 815, responder station 40 has determined that it is addressed, has cleared its locked-bit 88, has generated ARBITRATION NUMBER 101, and has begun transmitting response lDR. Simultaneously, responder station 36 has determined that it has been addressed, has generated ARBITRATION NUMBER 111, and has begun transmitting response lDR. Also, at time 815, simultaneous transmissions collide on common medium 32. 
At time 820, commander station 10 at block 226 chooses node 702 having BRANCH=OOO and MASK=001. Responder station 40 is not addressed because ARBITRATION NUMBER (101) ANDed with MASK (001) yields 001 which is not equal to BRANCH (000). Similarly, responder station 36 is not addressed. Neither station responds. At time 826, time out at block 218 occurs. 
At time 830 and block 226, a third group of responder station addresses is chosen. From FIG. 12 of the '551 patent the appropriate group is specified by traversing the tree according to a search method. If a breadth first search were used, all nodes at the same level would be visited before testing at a deeper level. Hence, node 703 would be next. If a depth first search were used, search would proceed upward from node 702 (because it is a leaf) and then downward from the first node having an untested branch. Hence, up to node 701 and down to node 703. As a refinement to either method, node 703 can be skipped because a collision at node 701 and no response at node 702 implies a collision at node 703 without testing. A depth first search would now traverse from node 703 directly to node 706. A breadth first search would first consider nodes 704 and 705 and conclude not to visit them because each is a descendent from a leaf node. 
Having selected node 706 at time 830, commander station 10 broadcasts an 10 command with BRANCH=001 and MASK=011 at block 216. At time 835 responder, station 40 has determined that it is addressed and has begun transmitting response IDR. Simultaneously, responder station 36 determines it is not addressed and remains in state 314. At time 840, shown on FIGS. 13 and 14 of the '551 patent, commander station 10 has received the response from responder station 40 as a proper message, concluded that only one responder station responded, derived received ARBITRATION NUMBER (101), set BRANCH to received ARBITRATION NUMBER, set MASK to all 1's so that a responder station must match ARBITRATION NUMBER (101) in all bit positions to respond, and begins to perform blocks 232 through 244 in FIG. 10 of the '551 patent. At time 845, responder station 40 has determined that it is addressed, has decoded a read command, has set its locked-bit 88 in state 326, and has begun generating the read response in state 328. At time 850, commander station 10 has received the response as a proper message. Thus, commander station 10 has conducted a first two-party uninterrupted command-response scenario from time 840 to time 850 with one responder station. 
The search by commander station 10 for another responder station proceeds from block 244 to block 224 in FIG. 10 of the '551 patent. At block 226, another node from FIG. 12 of the '551 patent is selected. Having elicited a proper response at node 706, the depth first search proceeds up to the first node having an untested branch, here node 703. Then, down the untested branch to node 707. Having selected node 707 at time 850, commander station 10 broadcasts an 10 command with BRANCH=011 and MASK=011 at block 216. At time 855, responder station 36 has determined that it has been addressed and has begun generating an lOR response. At time 860, the response is received by commander station 10 as a proper message. After time 860, events proceed in a manner similar to events from time 840 to time 850, as commander station 10 conducts a second two-party uninterrupted command-response scenario with a second responder station. 
At block 224, following the uninterrupted scenario, commander station 10 can conclude that all groups have been tested. On a depth first search, a proper response or no response at a node having BRANCH=MASK indicates all groups have been tested. On a breadth first search, all groups have been tested when an investigation of all levels up to the level having all MASK bits set to ‘1’ yields no node that is not descendent from a leaf node. 
In a branch/mask embodiment of the type described above, a responder station concludes that it has been addressed when ARBITRATION NUMBER logically ANDed with MASK is equal to BRANCH. Two other types of embodiments will now be described that lie within the scope and spirit of the present invention as described in the '551 patent. First, in an example of a high/low embodiment, BRANCH and MASK (as shown in format 142) are replaced with HIGH LIMIT and LOW LIMIT. Each of these limit values has the same precision as the MASK value. Using these limit values, responder station 40 concludes that it is addressed when HIGH LIMIT is greater than or equal to ARBITRATION NUMBER, and LOW LIMIT is less than or equal to ARBITRATION NUMBER. Second, in an example of a limit/bound embodiment, BRANCH and MASK (as shown in format 142) are replaced with a single LIMIT value having the same precision as MASK. Using a value called BOUND which by design choice may be 0 or the maximum permitted by the precision of LIMIT, responder station 40 concludes that it is addressed when ARBITRATION NUMBER is between BOUND and LIMIT, inclusive of both BOUND and LIMIT values. 
An example of a limit/bound embodiment is implemented with a structure similar to that already described for the branch/mask embodiment. Subtraction capability or equivalent must be added to ALU 72. Operation of microsequencer 42 must be revised to perform the arithmetic operations outlined above in state 312 shown on FIG. 11 of the '551 patent. The high/low embodiment is implemented with the structure already described for the limit/bound embodiment. 
In FIG. 10 of the '551 patent (blocks 210 and 226) commander station 10 specifies a group of responder station addresses. For a branch/mask embodiment, a method using the binary tree of FIG. 12 of the '551 patent has already been discussed. For a high/low embodiment, a similar binary tree (not shown) with HIGH and LOW values at each node is used. At the root node, LOW is 0 and HIGH is the maximum value permitted by the precision of the value HIGH. At a node on the lower left from a parent node, the value of LOW is the value of LOW at the parent node and the value of HIGH is a value ½ the value of HIGH at the parent node discarding a remainder, if any. At a node on the lower right from a parent node, the value of HIGH is the value of HIGH at the parent node and the value of LOW is ½ the value of HIGH at the parent node plus one. Although a binary tree has been described, a tree having more than two branches at each node can be employed to practice the present invention as described in the '551 patent as is readily apparent to those skilled in the art. Trees with varying number of branches at each node can also be employed. Operation of the high/low embodiment is otherwise identical to operation of the branch/mask embodiment already discussed. 
In a limit/bound embodiment, the method used to specify a group of responder station addresses is similar to the method described for a high/low embodiment with a minor variation in the tree. When BOUND is zero, then the value for LOW is not used and the value for HIGH is used as the LIMIT value at each node. When BOUND is a maximum value, then the value for HIGH is not used and the value of LIMIT at each node is the value of LOW. Operation of a limit/bound embodiment is otherwise identical to operation of a branch/mask embodiment already discussed. Note that the command at block 232 on FIG. 10 of the '551 patent sets locked-bit 88 to prevent unnecessary collisions when an 10 command using LIMIT is broadcast subsequently at block 228. 
FIG. 15 of the '551 patent is a fibonacci tree diagram for use in an example of an embodiment of the type already described as limit/bound. An advantage of using a fibonacci tree is that the LIMIT value for a node descendent from a parent node can be derived without a multiplication or division operation. In systems where it is desirable to improve calculation speed or reduce the complexity of circuitry and software at commander station 10, the fibonacci tree is used. An implementation of a high/low embodiment using a fibonacci tree similar to FIG. 15 of the '551 patent is within the ordinary skill of the systems design and programming arts. 
As described in several embodiments above, a commander station can quickly determine the identity of all responder stations coupled to a common medium at a given time. After the identity of a responder station has been determined, a commander station can conduct uninterrupted communication at any subsequent time using the responder station's ARBITRATION NUMBER. Since the ARBITRATION NUMBER is not unique, there is some risk that at a subsequent time, more than one responder station having a given value for ARBITRATION NUMBER may become coupled to the common medium. For increased accuracy, use of a unique responder station identity, such as the TAG field in format 146 of FIG. 6 of the '551 patent, may be used for subsequent communication. 
When more than one commander station is coupled to a common medium, it is possible for one commander station to thwart the objective of a second commander station. For example, when commander station 10 is attempting to identify all responder stations and commander station 34 issues an 10CG command, commander station 10 may subsequently incorrectly conclude that all responder stations have been identified. Several methods of preventing this incorrect conclusion are available to those skilled in communication and computer programming arts. Exemplary methods include enabling a commander station to monitor commands issued by another commander station to avoid inappropriate conclusions; enabling a commander station 10 to record the TAG fields sent in messages to another commander station and communicate directly with each such responder station, perhaps prior to and so simplifying, the task of identifying all responder stations; modifying the communication protocol used among commander stations; and causing a second commander station to delay its own attempt to identify all responder stations until after a time sufficient for a first commander station to identify all responder stations. The latter suggestion is practical using the media access control scheme of the present invention as described in the '551 patent. It is practical because the time required to identify a worst-case population of responder stations can be predetermined. 
The present invention as described in the '551 patent can be implemented in several alternate embodiments. As already discussed, various alternatives are available for common medium 32 including all media that support forms of electromagnetic energy, all sound, vibration, and pressure wave conducting media, and all media capable of transporting variation in chemical concentration, to name a few. If a medium other than radio communication is selected as an embodiment of the present invention as described in the '551 patent, variations in network interface 26 and 60 can be made by those skilled in the arts that apply to the selected medium. Appropriate signal sensors and generators are well known in applications including measurement and control apparatus. Packet synchronization techniques, packet formats, error detection techniques, and error correction techniques may vary or be omitted as a matter of design choice depending on the need for receiver synchronization, the signal to noise properties of the selected media, and the desired simplicity of network interfaces. 
Another group of alternative embodiments uses various means to specify a set of responder station identities or designations. The embodiments described above employ an ARBITRATION NUMBER selected from a predetermined range of numbers and expressed as part of a message. For example, alternate sets of designations include a set of operating modes, a set of modulation techniques, and a range of values used to shift in time all or a portion of a command. Various alternatives are also available for specifying (i.e. addressing) a subset of designations. The branch/mask, high/low, and limit/bound subset addressing techniques can each be applied to one or more parametric quantities related to the above mentioned set designations. For example, if onemember of the set is characterized by a bandwidth, a channel frequencies, a phase variation, or a duration in time, then a range of each of these parameters could be described by a branch/mask pair of values. 
Various alternatives for transmitting the command signal are within the scope of the present invention as described in the '551 patent. In the embodiments described at length above, the BRANCH and MASK values in the message format characterize the transmitted command signal according to a subset of responder stations to which the command is directed. In addition to the variations in modulation already described, the transmitted signal can be characterized by variation in the spread spectrum chip sequence or initial code within a chip sequence when spread spectrum transmission is employed. 
Other characteristics of a command signal can be used to limit or expand the subset of responder stations to which the command is directed. For example, the operation of commands including RO and WR to set locked-bit 88 and the operation of commands including IDG and IDCG to conditionally clear locked-bit 88 show how the command opcode can be used to characterize a command signal according to a selected subset or address range. Alternatively, modulation variations, timing variations, or other message content variations could also be used to set or clear an equivalent of the locked-bit function. 
Various means are suitable for use by a responder station to determine whether it is addressed by, i.e. whether it is a member of the subset indicated by, a command signal. Several arithmetic comparison techniques based on message content have been described above. Other means, based on whether the signal received by the responder is received without error, are appropriate when variations in modulation are used to address a subset of responder stations. For example, received signal strength below a threshold over one or more frequencies or at a particular time could cause commands to be received or rejected. Similarly, operation of functions similar to locked-bit 88 as already described and variation in spread spectrum chip sequence could be used to cause commands to be received or rejected. 
Within the scope of the present invention as described in the '551 patent, each responder station includes means for establishing a self designation. In the embodiments discussed at length above, the self designation is determined by a random number generator, held in a register, and included in a response packet. Alternative techniques include various means for sampling a random process to acquire an analog parametric value and using either a digital or an analog value to control the functions of network interface 60. 
Network interface 60 can be constructed and operated in several alternative embodiments to transmit a response packet in a way characterized by the responder station self designation. All of the following variations could be used in embodiments that fall within the scope of the present invention as described in the '551 patent: variations in the modulation technique, including variation within a range of values used to shift in time all or a portion of a response; variation in the spread spectrum chip sequence or initial code within a chip sequence when spread spectrum transmission is employed; variation in message content including preamble, postamble, response type indicator e.g. IDR, RDR, and WDR, register contents, status and locked-bit information; and variation based on signal rejection including variation in bandwidth, channel frequencies, signal phase variations, signal duration, or variation in the redundancies used to detect or correct transmission error. 
Another group of alternative embodiments uses alternative means for selecting a subgroup in response to collision detection. The tree search method that was described as part of the commander station protocol can be implemented in various ways depending on the selected representation of the tree in commander station memory 18. Binary trees have been described above. Other tree structures including n-ary trees could be employed to perform the commander station identification function in an equivalent manner. Depending on the type of tree selected for representation, the use of strings, arrays, stacks, pointers, linked lists, or derivative memory organizations are feasible and equivalent. Finally, tree search methods include depth first, breadth first, and combinations of both depth and breadth searching. 
Each computer used as part of commander station 10 and as part of responder station 40 includes hardware and software designed to conduct the protocol shown and described in S. 10 and 11 respectively of the '551 patent. Variations in the extent and complexity of hardware and software are well known by designers of ordinary skill in communication and computer arts. Equivalent hardware can include the general purpose computer such as an IBM PC; a calculator, such as an HP21C; the special purpose computer, such as application specific automated controllers used in weighing systems; the microprocessor based system, such as a circuit using an Intel 8048; the microsequencer based system using programmable devices and logic devices; and the integrated circuit or chip set, such as developed from a cell library using semiconductor device design techniques. Variations in the extent and complexity of software compatible with one or more of the above mentioned hardware variations are also well known by the programmer of ordinary skill. 
The systems designer of ordinary skill chooses to implement each control function in either hardware or software or a combination of both. A computer is said to conclude a certain result when it has determined the state of a control function. When a control function is implemented using a computer system, variations in the form of the result of the control function are well known. For example, a parameter that results from a first control function and is relied upon by a second control function can take the form of a signal when the second control function is in part hardware or the form of a pointer, value, or symbol stored in a register or memory when the second control function is in part software. 
The present invention as described in the '551 patent has been described in the preferred embodiment. Several variations and modifications have also been described and suggested. Other embodiments, variations, and modifications known to those skilled in the art may be implemented without departing from the scope and spirit of the invention as recited in the claims below.

Claims

1. A method of testing the RF communication operation of an RF transponder, comprising the steps of:

providing a sheet characterized by first and second opposite faces and a thickness;

mounting on the sheet an RF transponder that includes a transponder RF antenna;

positioning a first RF shield so as to abut the first face of the sheet;

positioning a second RF shield so as to abut the second face of the sheet, the second RF shield being in the shape of a cup having a mouth abutting said second face, wherein the first and second RF shields are positioned so that the first and second RF shields together form a closed cavity which completely surrounds and encloses the transponder RF antenna except where the thickness of the sheet separates the first RF shield from the mouth of the second RF shield, wherein said thickness is sufficiently small so that the first and second RF shields prevent any RF signals within the cavity from radiating outside the cavity;

positioning a test fixture RF antenna within the cavity;

transmitting an RF signal from the test fixture antenna;

detecting a response by the transponder to the RF signal; and

subsequently removing the transponder from proximity to the first and second shields and the test fixture RF antenna, so that no shielding obstructs the transponder RF antenna from sending and receiving RF radiation at any angle.

2. A method according to claim 1, wherein the cavity encloses the entire RF transponder.

3. A method according to claim 1, wherein the sheet has no shielding mounted thereon that obstructs RF radiation from the transponder RF antenna.

4. A method according to claim 1, wherein:

the first RF shield is in the shape of a cup having a mouth abutting the first face; and

the step of positioning the second RF shield further comprises aligning the mouth of the second shield with the mouth of the first shield.

5. A method according to claim 1, wherein the step of positioning the test fixture RF antenna within the cavity comprises:

mounting the test fixture RF antenna to a surface of one of the two RF shields;

connecting an RF transmission line to the test fixture RF antenna; and

passing the transmission line through an opening in said one RF shield to extend outside the cavity.

6. A method according to claim 1, further comprising the step of:

fabricating the sheet to include electrically conductive material adjacent the mouth of the second RF shield so as to improve RF shielding of the cavity.

7. A method according to claim 1, wherein the RF signal is transmitted at a predetermined wavelength, and wherein the RF shields are dimensioned to improve the gain of the cavity at that wavelength.

8. A method according to claim 1, wherein the RF signal is transmitted at a predetermined wavelength, and wherein the RF shields are dimensioned so that the cavity resonates at that wavelength.

9. A method of testing the RF communication operation of a plurality of RF transponders, comprising the steps of:

providing a sheet characterized by first and second opposite faces and a thickness;

mounting on the sheet a plurality of RF transponders, wherein each transponder includes a transponder RF antenna;

positioning a first test fixture section having a first RF shield so that the first RF shield abuts the first face of the sheet;

positioning a second test fixture section so as to abut the second face of the sheet, wherein:

the second test fixture section includes a plurality of RF shields,

each RF shield in the second test fixture section is in the shape of a cup having a mouth abutting said second face of the sheet,

the first and second test fixture sections so that each RF shield in the second test fixture section encircles a corresponding one of the transponder RF antennas so as to form, in combination with the first RF shield, a closed cavity that completely surrounds and encloses said corresponding transponder RF antenna except where the thickness of the sheet separates the first RF shield from the mouth of said RF shield in the second test fixture section, and

said thickness is sufficiently small so that the first and second RF shields prevent any RF signals within the cavity from radiating outside the cavity;

positioning within each cavity a corresponding test fixture RF antenna;

transmitting an RF signal from each test fixture antenna;

detecting a response by each transponder to the RF signal transmitted by its corresponding test fixture antenna; and

subsequently removing each transponder from proximity to the first and second test fixture sections and the test fixture RF antennas, so that no shielding obstructs each transponder RF antenna from sending and receiving RF radiation at any angle.

10. A method according to claim 9, wherein the each cavity encloses the entire corresponding RF transponder.

11. A method according to claim 9, wherein:

the first RF shield is in the shape of a plurality of cups so that each cup has a mouth abutting the first face of the sheet; and

the step of positioning the second RF shield further comprises aligning each mouth of the second shield with a corresponding mouth of the first shield.

12. A test fixture for testing the RF communication operation of an RF transponder which is mounted on a sheet which extends beyond the perimeter of the transponder, the RF transponder having an antenna for receiving RF signals, comprising:

first and second RF shields, the second RF shield being in the shape of a cup having a mouth;

an alignment mechanism for positioning the first and second RF shields to abut opposite sides of the sheet so that the mouth encircles the transponder antenna and so that the combination of the first and second RF shields forms a closed cavity completely surrounding and enclosing the transponder antenna except where the sheet separates the two RF shields, wherein the distance by which the sheet separates the two RF shields is small enough to prevent any RF signals within the cavity from radiating outside the cavity; and

a test fixture RF antenna mounted within the cavity.

13. A test fixture according to claim 12, further comprising:

a test fixture RF transmitter having an output connected to the test fixture RF antenna so that the RF antenna radiates RF signals to the transponder RF antenna; and

a test fixture RF receiver having an input connected to the test fixture RF antenna so that the RF receiver receives any responses transmitted by the RF transponder in response to said RF signals.

14. A test fixture according to claim 12, wherein the cavity encloses the entire transponder.

15. A test fixture according to claim 12, wherein:

the first RF shield is in the shape of a cup having a mouth abutting the first face; and

the alignment mechanism aligns the mouth of the second shield with the mouth of the first shield.

16. A method according to claim 12, further comprising:

an RF transmission line connected to the test fixture RF antenna;

wherein the transmission line extends through an opening in one of the RF shields so as to extend outside the cavity.

17. A test fixture according to claim 12, further comprising a test fixture RF transmitter for providing to the transponder antenna RF test signals having a predetermined wavelength, wherein the first and second RF shields are dimensioned to improve the gain of the cavity at that wavelength.

18. A test fixture according to claim 12, further comprising a test fixture RF transmitter for providing to the transponder antenna RF test signals having a predetermined wavelength, wherein the first and second RF shields are dimensioned so that the cavity resonates at that wavelength.

19. A test fixture for testing the RF communication operation of a plurality of RF transponders mounted on a sheet, each RF transponder having an RF antenna, comprising:

a first test fixture section including a first RF shield;

a second test fixture section including a plurality of RF shields each of which is in the shape of a cup having a mouth;

an alignment mechanism for positioning the first and second test fixture sections to abut opposite sides of the sheet so that each RF shield in the second test fixture section encircles a corresponding one of the transponder antennas so as to form, in combination with the first RF shield, a closed cavity that completely surrounds and encloses said corresponding transponder RF antenna except where the sheet separates the first RF shield from the mouth of said RF shield in the second test fixture section, wherein the distance by which the sheet separates the first RF shield from each RF shield of the second test fixture section is small enough to prevent any RF signals within each cavity from radiating outside that cavity; and

a test fixture RF antenna mounted within each cavity.

20. A test fixture according to claim 19, wherein:

the first RF shield is in the shape of a plurality of cups so that each cup has a mouth abutting the sheet; and

the alignment mechanism aligns each mouth of the second shield with a corresponding mouth of the first shield.

21. An interrogator for performing radio frequency communications with a radio frequency identification (rfid) tag, the interrogator comprising:

an antenna:

a radio frequency transmitter communicatively coupled to the antenna and configured to transmit commands at a first radio frequency band, a first command and a second command including a selection indicator, the selection indicator identifying one or more of a plurality of possible rfid tags;

a radio frequency receiver communicatively coupled to the antenna and configured to receive one or more responses from a selected rfid tag at a second radio frequency band, the second radio frequency band being different than the first radio frequency band, at least one of the responses including a random number generated by the selected rfid tag, the selected rfid tag being identified at least in part by the selection indicator; and

a processor communicatively coupled to the radio frequency receiver and the radio frequency transmitter and configured to format commands and process responses to identify rfid tags, the processor using the random number to identify the selected rfid tag in at least one subsequent command.

22. The interrogator of claim 21, wherein the processor is further configured to select the first radio frequency band in accordance with a frequency-hopping algorithm.

23. The interrogator of claim 21, further comprising receiving from the selected rfid tag data associated with one or more memory locations contained on the selected rfid tag.

24. The interrogator of claim 21, wherein the interrogator identifies the selected rfid tag in subsequent communications using at least one random number provided to the interrogator by the selected rfid tag.

25. The interrogator of claim 21, further comprising transmitting a write command, the write command including identifying the selected rfid tag by at least one random number provided to the interrogator by the selected rfid tag.

26. The interrogator of claim 21, wherein the processor is further configured to set the selection indicator to identify all possible rfid tags.