HK1055039B - An apparatus for retransmission of signals in a communication system - Google Patents

Description

Apparatus for retransmitting signal in communication system

Technical Field

The present invention relates to communications. More particularly, the present invention relates to an apparatus for fast retransmission of signals in a communication system.

Background

In a communication system, the communication channel through which signals propagate between transmitting and receiving terminals is susceptible to various factors of the communication channel, causing changes in the characteristics of the communication channel. In a wireless communication system, these factors include, but are not limited to: fading, noise, interference from other terminals, etc. Thus, despite the extensive error control coding used, some packets are lost or received incorrectly at the receiving terminal. Unless defined differently, a packet refers to a signal unit that includes a preamble, a payload, and a quality metric. Therefore, automatic repeat request (ARQ) schemes are often used at the link layer of a communication system to detect lost or erroneously received packets at the receiving terminal and to request retransmission of these packets at the transmitting terminal. One example of ARQ is the Radio Link Protocol (RLP). RLP is a type of error control protocol known in the art as a NAK-based ARQ protocol. In a system known as "DATA SERVICE OptionsFrost streaming speech SYSTEMS: such an RLP IS described in TIA/EIA/IS-707-A.8 of RADIO LINK PROTOCOL TYPE 2 ", herein designated RLP2, and incorporated herein by reference.

Existing ARQ schemes implement retransmission of lost or erroneously received packets by using a sequence number that is unique to each packet. When a receiving terminal detects a packet having a sequence number higher than an expected sequence number, the receiving terminal declares that a packet having a sequence number between the expected sequence number and the sequence number of the detected packet is lost or erroneously received. The receiving terminal then sends a control message to the transmitting terminal requesting retransmission of the lost packet. Alternatively, if the transmitting terminal does not receive an acknowledgement from the receiving terminal, the transmitting terminal may retransmit the packet after some time out interval.

Therefore, existing ARQ schemes cause a large delay between the first transmission and the subsequent retransmission of a packet. The ARQ does not declare a lost or erroneously received frame until a next packet containing a sequence number higher than the expected sequence number is received, or until the time out interval expires. This delay results in a large variance in the end-to-end delay statistics, which has a further detrimental effect on the network throughput. Transport layer protocols such as the Transmission Control Protocol (TCP) implement a congestion control mechanism that reduces the number of waiting packets in the network based on the variance of the round-trip delay estimate. In fact, a large variance in delay results in a reduction in the traffic entering the network and subsequently in a reduction in the throughput of the communication system.

One way to reduce the delay and the variance of the delay is to avoid retransmissions by ensuring that the first transmission is correctly received with a higher probability. However, this approach requires a larger amount of power, which in turn reduces throughput.

In light of the above, there is a need in the art for an ARQ mechanism with low retransmission delay.

Disclosure of Invention

The present invention is directed to a method and apparatus for fast retransmission (QARQ) of signals in a communication system.

According to an aspect of the invention, a receiving terminal determines a quality metric for a packet of a received signal. The receiving terminal then sends a Short Acknowledgement (SA) to the transmitting terminal based on the quality metric of the packet. If the quality metric indicates that the packet was received in error, then the SA is referred to as a Negative Acknowledgement (NAK); otherwise the SA is referred to as an Acknowledgement (ACK) or acknowledgement.

In another aspect of the invention, there is a determinable relationship before a particular packet and SA; thus, the SA need not contain an explicit indication of which packet to retransmit.

According to another aspect of the invention, the SA is an energy bit.

According to another aspect of the invention, the transmitting terminal attempts retransmission of the packet a predetermined number of times.

According to yet another aspect of the present invention, conventional sequence number based ARQ is used in conjunction with the QARQ scheme.

The present invention provides an apparatus configured to retransmit a signal in a communication system, comprising: a preamble detector configured to detect and decode a preamble of a unit of the received signal from the preamble detector, the preamble detector configured to prevent decoding of the unit of the received signal if the preamble indicates that the unit of the signal is not to be decoded; a decoder configured to decode the content of a unit of the received signal; a 1 st processor configured to determine a quality metric of a unit of the signal; a 1 st feedback signal generator coupled to the 1 st processor and configured to generate a 1 st feedback signal in response to a quality metric of a unit of the signal; a 2 nd feedback signal generator for generating a 2 nd feedback signal according to a result of decoding a content of the unit of the received signal; and a 2 nd processor configured to instruct the 2 nd feedback signal generator to generate a 2 nd feedback signal according to a sequence number of a unit of the signal when the retransmission of the signal according to the quality metric is declared to fail.

The present invention also provides an apparatus for retransmitting a signal in a communication system, comprising: a data queue for storing a plurality of units of signals to be transmitted; an allocator for determining whether the unit of the signal is a new unit or a retransmission unit; a scheduler for scheduling a unit for transmitting the signal to a destination receiving terminal; a 1 st detector for detecting a 1 st feedback signal received from the destination receiving terminal; and a 2 nd detector for detecting a retransmission request; a retransmission processor configured to receive information about the unit requesting retransmission from the allocator, and configured to prepare the unit for retransmission according to the retransmission request; and a 1 st control processor coupled to the scheduler, configured to receive the 1 st feedback signal, to select a signal transmitted at a time instant that precedes a time instant at which the 1 st feedback signal is received by a sum of a round trip delay and a determinable delay, and to schedule the signal for retransmission, wherein the determinable delay is a difference between a 1 st time instant at which the retransmission request is transmitted and a 2 nd time instant, the 2 nd time instant being selected from a group comprising: a time instant when a unit of the signal is received; a time instant when the determination of whether to demodulate the unit of the signal is made; a time instant when a unit of the signal is demodulated; and a time instant when the quality measure is calculated.

Drawings

The features, objects, and advantages of the present invention will become more apparent from the detailed description set forth below when taken in conjunction with the drawings in which like reference characters identify correspondingly throughout and wherein:

fig. 1 is a block diagram of an exemplary communication system.

Fig. 2 is an illustration of an exemplary forward link signal structure.

Fig. 3 is a flow chart of an exemplary method of data processing at a transmitting terminal.

Fig. 4 is a flow chart of an exemplary method of data processing at a receiving terminal.

Fig. 5 is a detailed block diagram of the communication system of fig. 1.

Fig. 6 is a diagram illustrating timing associated with packet processing at a receiving terminal according to an embodiment of the present invention.

Detailed Description

Fig. 1 illustrates an exemplary communication system 100 capable of implementing embodiments of the present invention. The 1 st terminal 104 transmits signals to the 2 nd terminal 106 over a forward link 108a and receives signals from the 2 nd terminal 106 over a reverse link 108 b. The communication system 100 may be operated bi-directionally, with the terminals 104 and 106 operating as either a transmitter unit or a receiver unit, or both, depending on whether data is being transmitted or received at the terminals. In an embodiment of the wireless cellular communication system, the 1 st terminal 104 may be a Base Station (BS) and the 2 nd terminal 106 may be a Mobile Station (MS) such as a telephone, a laptop computer, a personal digital assistant, etc. The forward and reverse links may be of the electromagnetic spectrum.

A link typically includes a set of channels that carry logically different types of information. The channels may be transmitted according to a Time Division Multiplexing (TDM) scheme, a Code Division Multiplexing (CDM) scheme, or a combination of both. In the TDM scheme, these channels are distinguished in the time domain. The forward link is comprised of time slots in a periodic sequence of time intervals in which the channel is transmitted. Thus, one channel is transmitted at a time. In the code division scheme, these channels are distinguished by pseudo-random orthogonal sequences; thus transmitting the channels simultaneously. A code division scheme is disclosed IN U.S. Pat. No. 5,103,459, entitled "SYSTEM AND METHOD DFOR GENERATING SIGNAL WAVEFORMS IN CDMA CELLULAR TELEPHONE SYSTEM", assigned to the assignee of the present invention and incorporated herein by reference.

In one embodiment of the invention, the forward link includes a set of channels, such as a pilot channel, a medium access channel, a traffic channel, and a control channel. The control channel is a channel that carries signals to be received by all MSs monitoring the forward link. In one embodiment of the invention, data carried on the traffic channel, including first transmissions and fast retransmissions, may be demodulated without information provided on the control channel. In another embodiment, the control channel may carry information needed for demodulation of data carried on the traffic channel. For the forward link signal structure of the exemplary embodiment of the present invention, reference is made to fig. 2.

In one embodiment of the invention, the reverse link includes a set of channels, such as traffic channels and access channels. Reverse traffic channels are dedicated to transmissions from a single MS to the BSs making up the network. A reverse access channel is used by the MS to communicate with a BS in a network when the MS does not have a traffic channel.

For simplicity, communication system 100 is shown as including only one BS 104 and one MS 106. However, other variations and configurations of the communication system 100 are possible. For example, in a multi-user, multiple-access communication system, a single BS may be used to transmit data to several MSs simultaneously. Furthermore, the MS may receive transmissions from several BSs simultaneously IN a manner similar to SOFT HANDOFF, which is disclosed IN U.S. patent No. 5,101,501, entitled "SOFT HANDOFF IN a CDMA cell HANDOFF SYSTEM," assigned to the assignee of the present invention and incorporated herein by reference. The communication system of the embodiments described herein may include any number of BSs and MSs. Thus, each of the plurality of BSs is connected to a Base Station Controller (BSC)102 over a backhaul similar to backhaul 110. The backhaul 110 may be implemented with several connection types including, for example, microwave or wired E1 or T1, or fiber optics. Connection 112 connects wireless communication system 100 to a Public Switched Data Network (PSDN), not shown.

In an exemplary embodiment, each MS monitors a signal quality metric of the received signal from the BS. An MS receiving forward link signals from multiple BSs (e.g., MS 106) identifies the BS (e.g., BS 104) associated with the highest quality forward link signal. The MS 106 then generates a prediction of the data rate at which the Packet Error Rate (PER) of packets received from the selected BS 104 will not exceed the target PER. An exemplary embodiment uses a target PER of approximately 2%. MS 106 then calculates a rate at which the "tail probability" is greater than or equal to the target PER. The tail probability is the probability that the actual signal quality during the packet transmission period is less than the signal quality required to successfully decode the packet correctly at a given rate. MS 106 then explicitly sends a message to the selected BS 104 on the reverse link requesting a data rate at which the particular selected base station may send forward link data to MS 106.

In one embodiment of the invention, the message is sent on a data rate control channel (DRC). DRCs are disclosed in pending application Ser. No. 08/963,386 entitled "A METHOD AND ANAPPARATUS FOR HIGH RATE DATA TRANSMISSION," assigned to the assignee of the present invention AND incorporated herein by reference.

In another embodiment of the invention, a dedicated reverse link medium access channel (R-MACCH) is used. The R-MACCH carries DRC information, Reverse Rate Indicator (RRI), and SA information.

In an exemplary embodiment, the BS 104 monitors reverse channels from one or more MSs and transmits data on the forward link to only one destination MS during each forward link transmit time slot. BS 104 selects a destination MS (e.g., MS 106) based on a scheduling procedure designed to balance the grade of service (GoS) requirements of each MS with the requirements to maximize the throughput of system 100. In the exemplary embodiment, BS 104 conveys data to destination MS 106 only at the rate indicated by the most recently received message from the destination MS. This limitation eliminates the need for the destination MS 106 to perform rate detection on the forward link signal. MS 106 need only determine whether it is the intended destination MS during a given time slot.

In the exemplary embodiment, the BS transmits a preamble in slot 1 of each new forward link packet. The preamble identifies the intended destination MS. Once the destination MS establishes that it is the intended data receiving station in the time slot, the MS decodes the data in the associated time slot. In an exemplary embodiment, the destination MS 106 determines the data rate of the data in the forward link based on a request message sent by the MS 106. The number of forward link time slots used to transmit packets varies depending on the data rate at which the packets are transmitted. Packets transmitted at a lower rate are transmitted using a greater number of slots.

Once MS 106 determines that data is being sent for that MS 106, MS 106 decodes the packet and estimates a quality metric of the received packet. The quality metric of a packet is defined by a formula based on the contents of the packet, e.g., parity bits, Cyclic Redundancy Check (CRC), etc. In one embodiment of the invention, the quality metric is a CRC. The estimated quality metric and the quality metric contained in the received packet are compared and an appropriate SA is generated based on the comparison. As discussed with reference to fig. 5, the SA in the exemplary embodiment may include only 1 bit.

In one embodiment, the SA is ACK-based, i.e., an ACK message is sent from the MS to the BS if the packet is decoded correctly, and no message is sent if the packet is decoded in error.

In another embodiment, the SA is NAK-based, i.e., a NAK message is sent from the MS to the BS if the packet is decoded in error, and no message is sent if the packet is decoded correctly. One advantage of this approach is that high reliability and noise interference to other reverse links and power savings at the MS can be achieved. As discussed, since the BS is transmitting packets that are transmitted for only one MS, at most that MS transmits a NAK, low interference on the reverse link is achieved. In well-designed systems, the probability of the MS decoding the packet incorrectly is low. Further, if a NAK is a bit of 0 energy, the NAK contains lower energy. Thus, the MS is guaranteed to be able to allocate a large amount of power to infrequent transmission of NAK bits.

In yet another embodiment, ACK is a 1 st energy value and NAK is a 2 nd energy value.

The SA is then transmitted to the BS 104 on a channel on the reverse link 108 b. In one embodiment of the invention, the reverse link channel is a DRC.

In another embodiment of the present invention, a code channel orthogonal to the reverse link may be advantageously utilized. Since the BS is transmitting packets that are transmitted for only one MS, at most that MS transmits the SA, low interference on the reverse link is achieved. In well-designed systems, the probability of the MS decoding the packet incorrectly is low. Further, if SA is ACK as 0 energy bits or NAK as 0 energy bits, the orthogonal channel contains low energy. Therefore, the MS can allocate a large amount of power to infrequent transmission of the SA bits, ensuring high reliability and low interference to the reverse link.

In yet another embodiment of the present invention, a dedicated reverse link medium access channel (R-MACCH) is utilized. The R-MACCH carries DRC, RRI, and ACK/NAK information.

The BS 104 detects the SA and determines whether retransmission of the packet is necessary. If the SA indicates that retransmission is necessary, the packet is scheduled for retransmission, otherwise the packet is discarded.

In an exemplary embodiment, the QARQ scheme described above operates in conjunction with RLP, as will be disclosed in the following description.

Fig. 2 shows a forward link signal structure transmitted by each base station in an exemplary high data rate system. The forward link signal is divided into time slots of fixed duration. In an exemplary embodiment, each slot is 1.67 milliseconds long. Each time slot 202 is divided into two half-time slots 204 and a pilot burst 208 is transmitted in each half-time slot 204. In an exemplary embodiment, each slot is 2048 symbols long, corresponding to a slot length of 1.67 milliseconds. In the exemplary embodiment, each pilot burst 208 is 96 symbols long and is centered at the midpoint of its associated half-time slot 204. A reverse link power control (RPC) signal 206 is sent to either side of the pilot burst in each half-2 slot 204 b. In the exemplary embodiment, the RPC signal is transmitted in 64 symbols before and 64 symbols after the 2 nd pilot burst 208b of each slot 202 and is used to adjust the power of the reverse link signal transmitted by each subscriber station. In the exemplary embodiment, forward link traffic channel data is transmitted in the remaining portion 210 of half 1 slot and the remaining portion 212 of half 2 slot. In the exemplary embodiment, preamble 214 is 64 symbols long and is sent with each packet. Since the traffic channel stream is sent for a particular MS, the preamble is specific to the MS.

In the exemplary embodiment, the control channels are transmitted at a fixed rate of 76.8kbps and are time division multiplexed on the forward link. The preamble of the control channel is recognizable by all MSs as the control channel message is directed to all MSs.

Fig. 3 is an exemplary flow chart of a method by which a BS transmits or retransmits a packet to an MS using a QARQ. At step 300, the BS receives a payload unit for transmission to the MS.

At step 302, the BS determines whether the payload unit is a payload unit to be transmitted or a payload unit to be retransmitted. At this step, retransmission requests may only be originated by the RLP, as discussed with reference to fig. 1.

If the payload unit is to be transmitted, the method proceeds to step 304, where the payload unit is provided to the 1 st time queue.

If the payload unit is to be retransmitted, the method proceeds to step 306, where the payload unit is provided to the 1 st time queue.

At step 308, the BS assembles the payload units generated for a particular MS into packets whose structure is determined according to the transmission data rate. The data rate at which the packets are transmitted is based on a feedback signal received on the reverse link from the destination MS. If the data rate is small, a packet of data is sent in multiple forward link slots (referred to as a multi-slot packet). In an exemplary embodiment, the preamble is sent in a new packet. The preamble allows identification of the intended destination MS during decoding. In an exemplary embodiment, only the 1 st slot of the multi-slot packet is transmitted with the preamble. Alternatively, the preamble may be transmitted in each forward link time slot.

At step 310, the BS transmits packets according to a scheduling order as discussed with reference to fig. 1.

After the packet has been transmitted, the BS tests whether an SA corresponding to the transmitted packet is received at step 312. As disclosed with reference to fig. 6, the BS knows when to expect an SA.

If an ACK is received in the expected time slot (or no NAK is received), the method continues at step 314. At step 314, the packet is removed from the time 1 and retransmission queue and discarded.

If a NAK is received (or no ACK is received) in the expected time slot, the method continues at step 316. At step 316, the parameters controlling the retransmission are tested. The parameters ensure that a packet is not retransmitted repeatedly, thereby increasing buffer requirements and reducing the throughput of the communication system. In one embodiment, the parameters include, for example, the maximum number of times a packet can be retransmitted and the maximum time a packet can be held in the time 1 queue after a packet has been sent. If the parameters are exceeded, the packet is removed from the time 1 and retransmission queue and discarded at step 318. In this case, the QARQ retransmission process ends and the packet may be retransmitted upon receipt of a request from the RLP processor as discussed with reference to fig. 6. If the parameter is not exceeded, the packet is rescheduled for retransmission at step 320.

Fig. 4 is an exemplary flow chart of a method for an MS to generate a response to a BS using a QARQ. At step 400, the MS receives a packet from the BS.

At step 402, a preamble of the packet is extracted. The preamble is compared to a reference preamble at step 404. If the preamble indicates that the packet is intended for another MS at step 406, the packet is discarded at step 408 and flow returns to step 400 to wait for another packet. If the preamble indicates that the packet is intended for the MS, the MS decodes the packet and estimates a quality metric of the received packet at step 408.

At step 410, the estimated instruction metric is compared to the quality metric contained in the received packet. If the estimated quality metric does not match the quality metric contained in the received packet, the appropriate SA is transmitted at step 412. In an exemplary embodiment, the SA is a NAK represented by a non-0 energy bit. At step 414, a timer is started for the SA sent. The purpose of the timer is to limit the period of time for which the MS waits for retransmission of the payload unit of the erroneously decoded packet. In an exemplary embodiment, if the payload element of an erroneously decoded packet is not received within the timer expiration period of the NAK associated with the erroneously decoded packet, the QARQ processing is interrupted and the RLP processes the missing payload element. See steps 416-432 and the accompanying description.

If the packet is decoded correctly at step 410, the appropriate SA is sent at step 416. In an exemplary embodiment, the SA is an energy-free bit. The payload units contained in the packet are then stored in a buffer at step 418.

At step 420, the RLP sequence numbers of the payload units are tested against the expected values of RLP sequence numbers.

If the RLP sequence numbers indicate adjacency, it means that all payload units of packets sent to the MS are properly received. Accordingly, payload units with consecutive sequence numbers contained in the buffer are provided to the RLP layer at step 420.

If the RLP sequence numbers indicate no adjacency, then a timer corresponding to the last NAK sent (which started at step 414) is checked at step 422. If the timer has not expired, the MS waits for retransmission of the missing payload unit or expiration of the timer for the last NAK sent.

If the timer for a particular NAK, and thus a particular set of missing payload elements, expires, the QARQ scheme for those payload elements is interrupted. All payload units stored in the buffer having sequence numbers higher than the missing payload unit associated with a particular NAK and lower than the missing unit associated with the next NAK (if any) are provided to the RLP layer at step 424.

At step 426, the RLP layer checks the sequence number of the passed payload unit. If the sequence numbers indicate adjacency, the RLP layer transfers data from the buffer to the sink at step 428. Otherwise, at step 430, the RLP layer generates an RLP message requesting retransmission of the missing unit. In one embodiment of the invention, the RLP message requests retransmission of all lost units in the buffer. In another embodiment, the message only requests retransmission of the recently detected missing payload units.

The message is transmitted to the serving BS on the reverse link at step 432.

Fig. 5 shows a detailed block diagram of the communication system 100 of fig. 1. Data to be delivered to MS 106 arrives at BSC 102 from a PSDN (not shown) via connection 112. The data is formatted into payload units under the control of the RLP processor 504. Although an RLP processor is shown in the embodiment, other protocols that allow retransmission according to the sequence number method may be utilized. In one embodiment of the invention, the payload unit is 1024 bits long. The RLP processor 504 also provides the allocator 502 with information regarding which packet has been requested for retransmission. The retransmission request is passed to the RLP processor 504 via RLP messages. The allocator 502 distributes the payload elements over the backhaul to the BS serving the MS for which the data is intended. The dispatcher 502 receives information about the location of the MS from the BS that serves the MS over the backhaul.

Payload units arriving at the BS 104 over the backhaul 110 are provided to a distributor 506. The allocator 506 tests whether the payload unit is a new payload unit or a payload unit provided by the RLP processor 504 for retransmission. If the payload unit is to be retransmitted, the payload unit is provided to a retransmission queue 510. Otherwise, the payload unit is provided to time 1 queue 508. The payload units are then assembled into packets according to the data rate requested by the MS 106, as described with reference to fig. 1.

The assembled packets are provided to the scheduler 512. The scheduler 512, in cooperation with the QARQ controller 518, assigns priorities between the time 1 packet and the packet for retransmission to the MS 106. Packets sent to the MS 106 remain in the queues 508, 510 while the BS 104 waits for the SA from the MS 106.

A packet arriving at the MS 106 on the forward link 108a is provided to a preamble detector 520, which detects and decodes the preamble of the packet. The preamble is provided to a processor 521 which compares the decoded preamble with a reference preamble. Discarding the packet if the preamble indicates that the packet is sent for another MS; otherwise the packet is provided to a decoder 522 which decodes the packet. The decoded packet is provided to a processor 521 which estimates a quality metric of the packet. The estimated quality metric is compared to the quality metric contained in the received packet and the appropriate SA is generated by the SA generator 526 based on the result of the comparison. Although the preamble detector 520, the decoder 522, and the processor 521 are shown as separate elements, those of ordinary skill in the art will appreciate that the physical distinction is made for illustrative purposes only. The preamble detector 520, the decoder 522, and the processor 521 may be combined into a single processor that implements the above-described processing.

If the packet is decoded in error, i.e., the estimated quality metric does not match the quality metric contained in the received packet, the SA is sent and a timer 530 for the SA is started. In an exemplary embodiment, the SA is a NAK represented by a non-0 energy bit. The purpose of the timer 530 is to limit the time period for the MS 106 to wait for retransmission of the payload unit of the erroneously decoded packet. The QARQ process is interrupted if the payload unit of the erroneously decoded packet is not received within the expiration period of the NAK's timer 530 associated with the erroneously decoded packet. Retransmission of lost payload units is handled by the RLP.

If the packet is decoded correctly, the payload units contained in the packet are stored in buffer 528. The RLP sequence numbers of the payload units contained in the packets are checked by the decoder 522 against the expected values of the RLP sequence numbers. If the RLP sequence numbers indicate adjacency, all payload units with consecutive sequence numbers contained in buffer 528 are provided to RLP processor 526. Otherwise, the timer corresponding to the last NAK sent is checked 530. If the timer has not expired, the payload units are stored in buffer 528 and MS 106 waits for the retransmission of the missing payload unit or expiration of timer 530 for the last NAK sent. If timer 530 expires for a particular NAK, and therefore a particular set of missing payload units, all payload units in buffer 528 having sequence numbers higher than the missing payload unit associated with the particular NAK and lower than the missing unit associated with the next NAK (if any) are provided to RLP processor 526.

RLP processor 526 checks the sequence number of the passed payload unit. If the sequence number indicates adjacency, the RLP processor 524 passes data from the buffer 528 to the data sink 534. Otherwise, the RLP processor 526 instructs the RLP message generator 532 to generate an RLP message requesting retransmission of the missing unit. In one embodiment of the invention, the RLP message requests retransmission of all lost units in buffer 528. In another embodiment, the message only requests retransmission of the recently detected missing payload units. The message is then transmitted to the BS 104 on the reverse link 108 b.

Data containing the SA and arriving at the BS 104 on the reverse link is provided to an SA detector 514 and an RLP message detector 516.

If the received data contains an ACK detected in the SA detector 514, the QARQ controller 518 removes the packet associated with the ACK from the queues 508, 510.

If a NAK is received, the QARQ controller 518 checks if the parameters controlling retransmission are exceeded. In an exemplary embodiment, the parameters include a maximum number of times a packet may be retransmitted and a maximum time a packet can be held in the time 1 queue 508 after a packet has been sent. If the parameter is exceeded, the QARQ controller 518 removes the packet from the queues 508 and 510. Otherwise, the QARQ controller 518 instructs the scheduler 512 to reschedule the packet for transmission at a higher priority. If the QARQ controller 518 determines that the negatively acknowledged packet resides in the 1 st time queue 510, the packet is moved from the 1 st time queue 508 to a retransmit queue 510.

If the received data contains an RLP retransmission request detected by an RLP message detector 516, the detector 516 provides an RLP message to the RLP processor 504 over the backhaul 110. The RLP processor then begins the process of retransmitting packets according to the implemented RLP.

Fig. 6 illustrates the relationship between packets received at the MS 106 and SAs transmitted from the MS 106. In time slots n-4, n-3, a receiver at MS 106 receives a packet on forward channel link 108 and determines whether the packet originated for that MS 106. If the packet is not addressed to the MS 106, the MS 106 discards the packet. Otherwise, the MS 106 decodes the packet, estimates a quality metric for the packet, and compares the estimated quality metric to the quality metrics contained in the packets in slots n-2, n-1. In time slot n, the transmitter at the MS 106 sends the SA back to the BS 104 on the reverse channel link 108 b. In time slot n +1, the SA received at the BS 104 is decoded and provided to the QARQ controller. If requested, the BS 104 retransmits the packet in slots n +2, n + 3. The positions of the received time slots on the forward link channel 108a and the reverse link channel 108b are synchronized at the MS 106. Thus, the relative positions of the time slots on the forward link channel 108a and the reverse link channel 108b are fixed. The BS 104 can measure the round trip delay between the BS 104 and the MS 106. Thus, if the relationship between the received packet processing and the SA is determinable, the time slot in which the SA must arrive at the BS 104 can be determined.

In one embodiment of the invention, the relationship between the processing of the received packet and the SA is determined by the number of fixed slots between the received packet and the sending back of the SA (i.e., slots n-2, n-1). Thus, the BS 104 may associate each packet with each SA. One of ordinary skill in the art will appreciate that the meaning of fig. 5 is merely to illustrate the concept. Thus, the number of time slots allocated for an event may be changed, e.g., decoding and estimation of the quality metric of the packet may occur in more or less than two time slots. Furthermore, certain events are variable in nature, such as the length of the packet, the delay between the reception of the SA and the retransmission of the packet.

In another embodiment of the invention, the relationship between the processing of the received packet and the SA may be determined by including information about which packet is to be retransmitted to the SA.

The previous description of the preferred embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without the use of the inventive faculty. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (16)

1. An apparatus configured to retransmit a signal in a communication system, comprising:

a preamble detector configured to detect and decode a preamble of a unit of the received signal, the preamble detector configured to prevent decoding of the unit of the received signal if the preamble indicates that the unit of the signal is not to be decoded;

a decoder configured to decode the content of units of the received signal from the preamble detector;

a 1 st processor configured to determine a quality metric of a unit of the signal;

a 1 st feedback signal generator coupled to the 1 st processor and configured to generate a 1 st feedback signal in response to a quality metric of a unit of the signal;

a 2 nd feedback signal generator for generating a 2 nd feedback signal according to a result of decoding a content of the unit of the received signal; and

a 2 nd processor configured to instruct the 2 nd feedback signal generator to generate a 2 nd feedback signal according to a sequence number of a unit of the signal when the retransmission of the signal according to the quality metric is declared to fail.

2. The apparatus of claim 1, wherein the 2 nd processor is further configured to declare the retransmission of the signal according to the quality metric as failed when:

a unit that does not receive the signal within a predetermined number of retransmissions; or

A unit that does not receive the signal within a predetermined time period measured from a first transmission of the unit of the signal; or

The unit of the signal is not received within a predetermined time period measured from transmission of a retransmission request corresponding to the unit of the signal.

3. The apparatus of claim 1, wherein the units of the signal are packets.

4. The apparatus of claim 1, wherein the quality metric is a cyclic redundancy check.

5. The apparatus of claim 1, wherein the decoder decodes the content of the units of the signal according to information carried on a control channel.

6. The apparatus of claim 1, wherein the 1 st feedback signal is an energy pulse.

7. The apparatus of claim 6, wherein the energy pulse is one bit.

8. The apparatus of claim 1, wherein the 1 st feedback signal contains no energy.

9. The apparatus of claim 8, wherein the 1 st feedback signal is a bit.

10. The apparatus of claim 1, wherein the 1 st processor is further configured to transmit the 1 st feedback signal at a determinable time instant, wherein the determinable time instant is fixedly delayed from an event time instant, the event time instant being selected from a group comprising:

a time instant when a unit of the signal is received;

a time instant when the determination of whether to demodulate the unit of the signal is made;

a time instant when a unit of the signal is demodulated; and

a time instant when the quality measure is calculated.

11. An apparatus for retransmitting signals in a communication system, comprising:

a data queue for storing a plurality of units of signals to be transmitted;

an allocator for determining whether the unit of the signal is a new unit or a retransmission unit;

a scheduler for scheduling a unit for transmitting the signal to a destination receiving terminal;

a 1 st detector for detecting a 1 st feedback signal received from the destination receiving terminal; and

a 2 nd detector for detecting a retransmission request;

a retransmission processor configured to receive information about the unit requesting retransmission from the allocator, and configured to prepare the unit for retransmission according to the retransmission request; and

a 1 st control processor coupled to the scheduler configured to receive the 1 st feedback signal, to select a signal to transmit at a time instant that is prior to a time instant at which the 1 st feedback signal is received by a sum of a round trip delay and a determinable delay, and to schedule the signal's unit for retransmission,

wherein the determinable delay is a difference between a 1 st time instant at which the retransmission request is transmitted and a 2 nd time instant, the 2 nd time instant being selected from a group consisting of:

a time instant when a unit of the signal is received;

a time instant when the determination of whether to demodulate the unit of the signal is made;

a time instant when a unit of the signal is demodulated; and

a time instant when the quality measure is calculated.

12. The apparatus of claim 11 wherein said determinable delay is included in said 1 st feedback signal.

13. The apparatus of claim 11, wherein the 1 st processor is further configured to determine a time instant at which to retransmit the unit of the signal, the time instant variably delayed from receipt of the 1 st feedback signal.

14. The apparatus of claim 11, wherein the 1 st processor is further configured to determine a time instant at which to retransmit the unit of the signal, the time instant fixedly delayed from receipt of the 1 st feedback signal.

15. The apparatus of claim 11, further comprising:

a 2 nd detector for detecting the 2 nd feedback signal received from the destination receiving terminal,

the retransmission processor is further configured to receive the 2 nd feedback signal, select a signal based on the 2 nd feedback signal, and prepare the signal for retransmission.

16. The apparatus of claim 15 wherein said 2 nd feedback signal contains a sequence number of a unit of said signal to be retransmitted.