WO2006106617A1

WO2006106617A1 - Ip packet mapping method

Info

Publication number: WO2006106617A1
Application number: PCT/JP2006/306114
Authority: WO
Inventors: Nobuhiko Miki; Hiroyuki Atarashi; Atsushi Harada; Naoto Okubo; Mamoru Sawahashi
Original assignee: Ntt Docomo, Inc.
Priority date: 2005-04-01
Filing date: 2006-03-27
Publication date: 2006-10-12
Also published as: TW200705928A; TWI307229B; JP2006311464A

Abstract

An IP packet mapping method includes: a step for generating MAC-PDU directly if an IP packet is smaller than a predetermined size or after dividing the IP packet into a plurality of parts if it exceeds the predetermined size; a step for adding a header containing the IP packet division information and a CRC code to the MAC-PDU; a step for channel-encoding the MAC-PDU having the header and the CRC code so as to generate a channel coding block; and a step for mapping the channel coding block to a radio frame.

Description

Specification

IP packet mapping method

Technical field

[0001] The present invention relates to an IP packet mapping method for mapping an IP (Internet Protocol) packet to a radio frame of a physical channel.

Background art

FIG. 1 is a diagram showing a method for mapping IP packets to physical channels in conventional HSDPA (High Speed Downlink Packet Access).

[0003] In FIG. 1, the IP packet to be transmitted is first divided into RLC—PDU (Protocol Data Unit) of a certain size, which is a retransmission unit of RLC (Radio Link Control) layer. (Step SI). A header H (RLC header) including control information such as a packet number for realizing retransmission in the RLC layer is added to the head of the RLC-PDU.

[0004] Next, RLC—PDUs that can be transmitted in one frame are collected according to the channel state, and a MAC (Media Access Control) —PDU is generated (step S2).

[0005] Next, a header MAC-H (MAC header) including control information in the MAC layer is attached to the head portion of the MAC-PDU, and an error for performing retransmission in the MAC layer A detection code CRC (Cyclic Redundancy Check) code is added (step S3).

[0006] Next, the MAC PDU to which the MAC header and CRC code are attached is subjected to channel coding for error correction, and becomes a channel coding block (Channel Coding Biock) (step S4). This channel coding block is an ARQ (Automatic Repeat reQuest) block which is a retransmission unit in the MAC layer.

[0007] Then, the channel coding block is mapped to a radio frame (Step S5) and transmitted.

[0008] The applicant has not been able to find prior art documents related to the present invention by the time of filing. Therefore, prior art document information is not disclosed.

Disclosure of the invention

Problems to be solved by the invention [0009] The conventional mapping method of IP packets to physical channels in HSDPA has been performed as described above, and has the following problems.

[0010] First, since the mapping is performed on the radio frame of the physical channel via a plurality of layers such as the RLC layer and the MAC layer, many headers are added along the way. There was a problem that communication efficiency could not be achieved due to overhead. Figure 2 shows an example of the number of bits such as headers in the mapping of a conventional IP packet to a physical channel. For example, if the IP packet is 1500 bytes and the RLC-PDU is 40 bytes, the IP packet is 38 bytes. This is divided into RLC-PDUs, each with a 16-bit RLC header. The header attached to the MAC-PDU is as shown in the figure. Because many headers are added in this way, the efficiency of communication cannot be achieved due to the overhead.

[0011] Second, although the unit whose quality should be finally maintained is an IP packet to be transmitted, retransmission for maintaining quality when a data error occurs is performed by a channel coding block (ARQ block). It is performed in two layers, the MAC layer with the unit of RLC and the higher RLC layer with the unit of RLC—PDU, and is not directly related to the IP packet. For example, if an IP packet is divided into 10 RLC-PDUs, even if 9 of them are successfully received, if an error occurs in one, the entire IP packet becomes meaningless and retransmitted. The control works effectively.

[0012] Thirdly, since IP packets have variable length, they can be mapped in various sizes, but in the conventional method, they are divided into fixed-size RLC—PDUs, and 1 depending on the channel conditions. RLC—PDUs that can be transmitted in a frame are grouped together to generate MAC—PDUs. Since the IP packet size is not taken into consideration here, it cannot be made sufficiently efficient. Figure 3 shows an example of the distribution of IP packet size on the wired Internet. (A) shows the ratio of the number of packets (vertical axis) to IP packet size (horizontal axis), and (b) shows IP packets. The cumulative value (vertical axis) of the ratio of the number of packets to the size (horizontal axis) of the packet tends to be very short (around 40 bytes) and large (around 1500 bytes) ( (Less than 64 bytes is about 50%, 1400 bytes and more is about 20%). In future wireless communications, IP packets will be the main traffic. It is desirable to be able to handle IP packets with a wide distribution size.

[0013] The present invention has been proposed in view of the above-described conventional problems, and its purpose is to reduce the overhead due to the header to improve efficiency and to make the retransmission unit closer to the IP packet. Therefore, it is an object of the present invention to provide an IP packet mapping method that can improve the performance of quality control and can appropriately cope with IP packets of a wide distribution size.

Means for solving the problem

[0014] In order to solve the above problems, according to the present invention, as described in claim 1, the IP packet is left as it is when it is smaller than a predetermined size, and is divided when it exceeds the predetermined size. To generate a MAC-PDU, add a header and CRC code including IP packet division information to the MAC-PDU, and send a MAC-PDU with the header and CRC code attached to the channel. A step of encoding to generate a channel coding block and a step of mapping the channel coding block to a radio frame are provided.

[0015] In addition, as described in claim 2, a step of generating one or a plurality of MA C PDUs having a fixed size from an IP packet, a header including IP packet division information in the MAC-PDU, and a CRC code A step of adding, a step of generating a channel coding block by channel-coding the MAC PDU to which the header and CRC code have been added, and a step of mapping the channel coding block to a radio frame. Can do.

[0016] Also, as described in claim 3, a step of generating a variable-size MAC-P DU as it is from an IP packet, and a header and a CRC code including IP packet division information in the MAC-PDU A step of performing an attached card, a step of generating a channel coding block by channel-coding the MAC-P DU with the header and CRC code attached thereto, and a step of mapping the channel coding block to a radio frame Can be provided.

[0017] In addition, as described in claim 4, in the IP packet mapping method according to claim 1, the predetermined size depends on reception quality information and Z or retransmission request signal frequency. It can be set adaptively according to this.

[0018] Also, as described in claim 5, in the IP packet mapping method according to claim 1, the division when the IP packet exceeds a predetermined size is performed by dividing the IP packet from the beginning. The predetermined size can be divided.

[0019] In addition, as described in claim 6, in the IP packet mapping method according to claim 1, in the case where the IP packet exceeds a predetermined size, the division is performed when the IP packet exceeds the predetermined size. It can be equally divided into a predetermined number so as to be equal to or smaller than the size of.

[0020] In addition, as described in claim 7, in the IP packet mapping method according to claim 1, a step of receiving reliability indicating the degree of error of the channel coding block from the receiving side, Depending on the reliability, a step of retransmitting a large number of bits when the reliability is low and a step of retransmitting only a part of the bits when the reliability is high can be provided.

[0021] In addition, as described in claim 8, according to the IP packet mapping method of claim 7, the reliability is received by a NACK signal divided into levels according to the reliability. It can be done.

[0022] Further, as described in claim 9, in the IP packet mapping method according to claim 1, the number of bits of the channel coding block from the receiving side is determined from the receiving side. A step of receiving the reliability indicated, and a step of retransmitting only the bit or block whose reliability is determined to be low and an error has occurred according to the reliability.

[0023] In addition, as described in claim 10, in the IP packet mapping method according to claim 1, the size of the IP packet as it is or divided is not equal to a predetermined value. Alternatively, if there is a margin in size, it is possible to perform concatenation with the following IP packet, rate matching by repetition code, or dummy bit insertion.

[0024] Also, as described in claim 11, in the IP packet mapping method according to claim 1, the retransmission request signal from the receiving side based on the CRC code is transmitted together with other control symbols as an ACKZNACK signal. It can be accompanied by another CRC code which is a CRC code. [0025] In addition, as described in claim 12, the IP packet is divided and connected to the Z or subsequent IP packet to generate a MAC-PDU, and to the MAC-PDU. A packet code part for adding a header and a CRC code including IP packet division information, and a channel code for generating a channel coding block by channel-coding a MAC-PDU with the header and CRC code attached thereto. It can be configured as a transmitter including a conversion unit and a mapping unit that maps the channel coding block to a radio frame.

The invention's effect

[0026] In the IP packet mapping method of the present invention, since the RLC layer header is eliminated, overhead can be reduced and efficiency can be improved. In addition, quality control performance can be improved by making the retransmission unit closer to the IP packet. Furthermore, by dividing and concatenating in several patterns according to the size of the IP packet, it is possible to efficiently transmit IP packets having a wide distribution size.

Brief Description of Drawings

[0027] FIG. 1 is a diagram showing a method for mapping IP packets to physical channels in conventional HSDPA.

FIG. 2 is a diagram showing an example of the number of bits such as a header in mapping of a conventional IP packet to a physical channel.

FIG. 3 is a diagram showing an example of IP packet size distribution on the wired Internet.

FIG. 4 is a flowchart showing an IP packet mapping method of the present invention.

FIG. 5 is a diagram showing pattern classification of the IP packet mapping method of the present invention.

[FIG. 6] A diagram showing an example of the relationship between an IP packet and a MAC-PDU when there is no division and variable length.

[Fig.7] IP packet and MAC in the case of splitting, variable length, equal splitting to be less than a certain length

— A diagram illustrating an example of a relationship of PDUs.

[Figure 8] IP packet and MAC with fragmentation, variable length, and fixed length (only one variable length)

— A diagram illustrating an example of a relationship of PDUs.

[Fig. 9] A diagram showing an example of the relationship between IP packets and MAC-PDUs when there is division and fixed length. The

FIG. 10 is a diagram showing a configuration example of a communication system to which the IP packet mapping method of the present invention is applied.

FIG. 11 is a diagram showing an example of generating a channel coding block from an IP packet via MAC-PDU.

FIG. 12 is a conceptual diagram of adaptive modulation / demodulation 'channel coding (AMC).

FIG. 13 is a diagram illustrating an example of a combination of a data modulation scheme and a channel code rate.

FIG. 14 is a diagram showing an example of mapping from channel coding blocks to radio frames.

FIG. 15 is a diagram illustrating hybrid ARQ processing.

[16] FIG. 16 is an explanatory diagram of a retransmission request signal error detection type hybrid ARQ based on a CRC code employed in an embodiment of the present invention.

[17] This is a diagram showing an example of a retransmission request signal error that occurs in a conventional general hybrid ARQ.

FIG. 18 is a diagram illustrating an example of compensation for a retransmission request signal error by a conventional RLC layer.

FIG. 19 is a diagram showing another configuration example of a communication system to which the IP packet mapping method of the present invention is applied.

FIG. 20 is a diagram illustrating an example of a NACK signal.

FIG. 21 is a diagram illustrating an example of an ACKZNACK signal that also indicates the overall reliability.

FIG. 22 is a diagram showing an example of a block.

FIG. 23 is a diagram showing a type of hybrid ARQ processing.

[24] FIG. 24 is a diagram illustrating a configuration example of a decoder.

[25] FIG. 25 is a diagram showing an example of detecting reliability using soft decision output.

[26] This is a diagram showing an example of detecting reliability using the number of code inversions in the iterative decoding process.

FIG. 27 is a diagram illustrating an example of a calculation process in a case where reliability is detected using cross-entropy in iterative decoding.

FIG. 28 is a diagram illustrating an example of detecting reliability using received SINR. Explanation of symbols

10 Transmitter

11 Transmission notifier

12 divisions, joints

13 Packet encoder

14 channel encoder

15 Mapping words

16 Modulator

17 Retransmission controller

18 Control unit

20 Receiver

21 Demodulator

22 Reception quality measurement unit

23 Retransmission controller

24 IP packet playback unit

25 Reliability detector

BEST MODE FOR CARRYING OUT THE INVENTION

[0029] Hereinafter, preferred embodiments of the present invention will be described.

FIG. 4 is a flowchart showing the IP packet mapping method of the present invention.

[0031] Fig. 4 (a) is a type that generates variable-length MAC PDUs without dividing the IP packet, and generates a MAC PDU as it is from the IP packet (Step S11). MAC—A step to add a header (including control information at the MAC layer) and CRC code (error detection code for retransmission at the MAC layer) including IP packet segmentation information to the PDU (step S12) And a step of generating a channel coding block by generating a channel coding block by error-correcting the MAC-PDU with the header and CRC code (step S13) and mapping the channel coding block to a radio frame (step S13). Step S14). Note that in the MAC-PDU generation process (step S11), if the IP packet size does not reach a predetermined value (channel coding When the packet is concatenated with the following IP packet (only when there is an IP packet following the transmission buffer), rate matching by repetition code 匕, or dummy bit insertion. And sometimes not. Further, these processes such as connection can be controlled according to the radio link state.

[0032] Fig. 4 (b) is a type that generates a variable-length MAC-PDU by dividing according to the size of the IP packet. If the IP packet is smaller than the predetermined size, it is left as it is. If it exceeds the maximum size, a step of generating a MAC-PDU by dividing (step S21) and a step of adding a header and a CRC code including IP packet division information to the generated MAC-PDU (step S22) And a MAC-PDU with a header and CRC code and a channel code to generate a channel coding block (step S23), a step of mapping the channel coding block to a radio frame (step S24), and It is configured. Note that in the MAC-PDU generation process (step S21), if the size of the IP packet as it is or the divided IP packet is less than a predetermined value, the rate of connection with the subsequent IP packet and the repetition code key There are cases where matching or dummy bit insertion is performed or not. In addition, the processing such as connection can be controlled according to the wireless link state.

[0033] Figure 4 (c) is a type that always generates a fixed-size MAC PDU, a process of generating one or more fixed-size MAC PDUs from an IP packet (step S31), and MAC—A step of adding a header and CRC code including IP packet segmentation information to the PDU (step S32), and a MAC—PDU with the header and CRC code attached to the channel PDU, and a channel coding block. The process includes a generating process (step S33) and a process of mapping channel coding blocks to radio frames (step S34). Note that in the MAC-PDU generation process (step S31), when the size of the generated MAC-PDU is smaller than a predetermined value, it is connected with the following IP packet, rate matching by repetition code key or dummy There are cases where bit insertion is performed and cases where bit insertion is not performed. In addition, these processes such as connection can be controlled according to the state of the radio link.

FIG. 5 is a diagram showing pattern classification of the IP packet mapping method of the present invention. P7 corresponds to the method of FIG. 4 (a), P8 to P20 correspond to the method of FIG. 4 (b), and P8 and P21 to P25 correspond to the method of FIG. 4 (c). In addition, P8 to P20 are further divided into P8 to P13 when equally dividing so as to be equal to or less than a certain length, and P8, P9, and P14 to P20 when dividing with a fixed length.

6 to 9 are diagrams showing examples of the relationship between the IP packet and the MAC-PDU in a representative one of the above patterns.

[0036] Fig. 6 is a diagram showing an example of the relationship between an IP packet and a MAC-PDU in the case of no division and variable length (corresponding to the methods P1 to P7 in Fig. 4 (a) and Fig. 5). This shows how the size of MAC-PDU is linked to the size of. If the size of the IP packet is less than a predetermined value, as shown by the notching and hatching (hatched area), it is connected with the following IP packet, rate matching by repetition code key もしくは or dummy bit insertion It also shows how this is done.

[0037] According to this method, there is an advantage that the overhead is minimized. In addition, since the unit of retransmission and channel code key is an IP packet whose final quality should be guaranteed, quality control is simplified. On the other hand, if the IP packet size is very large, such as 1500 bytes, the retransmission unit becomes very large, so the retransmission efficiency may be degraded. In addition, the size of the MAC-PDU varies greatly. As will be described later, the problem of efficiency deterioration due to retransmission can be solved by improving the retransmission mechanism.

[0038] Fig. 7 shows that there is a division, and the IP packet and MAC-PDU in the case of equal division (corresponding to the methods P8 to P13 in Fig. 4 (b) and Fig. 5) so that it is variable length and below a certain length. It is a figure showing an example of the relationship. Up to the threshold size S, the MAC-PDU size is linked to the IP packet size.

TH

However, when the threshold size S is exceeded, equal division is performed so that it is less than a certain length.

TH

It shows how it is called. If the size of the IP packet is less than a predetermined value, as shown by hatching, the connection with the subsequent IP packet, rate matching by repetition code key 匕, or dummy bit insertion is also performed Show me.

[0039] According to this method, a MAC-PDU having a very large size does not exist, and a force overhead for reducing variation in the size of the MAC-PDU is slightly increased. [0040] Figure 8 shows a fragmented, IP packet with variable length and fixed length (only one variable length) (corresponding to methods P8, P9, and P14 to P20 in Figure 4 (b) and Figure 5) Is a diagram showing an example of the relationship between MAC and PDU, and up to the threshold size S, the size of the IP packet

TH

When the size exceeds the threshold size S, the fixed length from the top

TH

The division is performed at, and the remaining part shows a variable length. If the size of the IP packet as it is or the divided IP packet does not reach a predetermined value, as shown by hatching, it is connected with the following IP packet, rate matching by repetition code key or dummy The state of inserting bits is also shown.

[0041] According to this method, there is no existence of a very large size MAC-PDU, and the MAC-PDU size variation is reduced. The overhead is slightly increased, and a small-size MAC-PDU is reduced. There is a possibility.

[0042] Fig. 9 is a diagram showing an example of the relationship between IP packets and MAC-PDUs in the case of division and fixed length (corresponding to the methods P8 and P21 to P25 in Fig. 4 (c) and Fig. 5). This shows how to generate a fixed size MA PDU at all times. If the size of the generated MAC-PDU is smaller than the predetermined value, connection with the following IP packet, rate matching by repetition code key, or dummy bit insertion is performed, as shown by (2) and (2). The situation is also shown.

[0043] According to this method, since the MAC-PDU is not larger than a certain size, mapping to a radio frame is facilitated. On the other hand, the overhead increases.

Next, FIG. 10 is a diagram showing a configuration example of a communication system to which the IP packet mapping method of the present invention is applied.

[0045] In FIG. 10, the transmitter 10 takes out the transmission buffer 11 that temporarily stores the IP packet to be transmitted, and the IP packet stored in the transmission buffer 11, and then splits and Z or continues the IP packet. A segmentation and concatenation unit 12 that performs concatenation with a packet and generates a MAC-PDU, a packet code input unit 13 that performs a header including an IP packet segmentation information and a CRC code attached to the generated MAC-PDU, and MAC with header and CRC code — Channel code unit 14 for generating channel coding blocks by channel encoding PDUs; Mapping unit 15 for mapping channel coding blocks to radio frames; Matsupi And a modulation unit 16 that modulates the converted radio frame into a radio signal. The transmitter 10 also includes a retransmission control unit 17 that performs retransmission control in accordance with a retransmission request signal transmitted from the receiver 20, and reception quality information and a retransmission request signal (frequency) transmitted from the receiver 20. A control unit 18 is provided to control the division size of the dividing / connecting part 12 and the presence / absence of connecting according to the situation. Here, when the reception quality is poor or when the frequency of retransmission request signals is high, the control unit 18 reduces the division size and performs control so that concatenation is not performed, and conversely, when the reception quality is good or retransmission is performed. If the frequency of the request signal is low, the division size is increased and control is performed in the direction in which concatenation is performed so that transmission can be performed more efficiently.

[0046] On the other hand, the receiver 20 includes a demodulator 21 that demodulates the radio signal, a reception quality measurement unit 22 that measures the reception quality from the demodulated signal and sends it to the transmitter 10, and the presence or absence of retransmission from the demodulated signal. And a retransmission control unit 23 that transmits a retransmission request signal to the transmitter 10 side, and an IP packet reproduction unit 24 that reproduces an IP packet from the demodulated signal.

FIG. 11 shows an example in which a channel coding block is generated from an IP packet via a MAC-PDU in division / concatenation unit 12, packet code key unit 13, and channel code key unit 14 in FIG. FIG. Figure 11 (a) shows the case where the IP packet is large. The IP packet is divided into multiple MAC-PDUs, and a header and CRC code are added to form a channel coding block. Figure 11 (b) shows the case where the IP packet is small, and shows a state where a plurality of IP packets are combined into one MAC-PDU and then header and CRC code are added to form a channel coding block. And

[0048] Unless the MAC-PDU size is linked to the IP packet size or fixed size, the control unit 18 (Fig. 10) controls the MAC-PDU size according to the communication status. Thus, adaptive modulation and channel coding (AMC) that appropriately controls the relationship between the data rate and the data error rate can be used. In this case, the size of the MAC-PDU varies depending on the channel condition, and the number of packets that can be transmitted in one frame changes. Fig. 12 is a conceptual diagram of adaptive modulation / demodulation 'channel coding (AM C). Assuming that transmission is performed with the same transmission power from the base station 100, the user # 1 in the vicinity of the base station 100 Since the terminal 201 has a large channel state with a large received power, a modulation scheme (for example, 16QAM (Quadratur e Amplitude Modulation)) and channel code ratio (large). In addition, the terminal 202 of the user # 2 located far away has a low received power and a poor channel state. Therefore, although the data error rate is low although the data rate is low, the modulation method (for example, QPSK (Quadrature Phase Shift Keying)) And select the channel coding rate (small). FIG. 13 is a diagram showing an example of a combination of a data modulation scheme and a channel code rate, in which the data rate increases but the data error rate increases in the direction of the arrow (downward). Therefore, the lower the data modulation scheme and the channel code rate combination are applied, the better the channel state, and the lower the data modulation scheme and channel code rate combination is applied. In practice, a table is prepared that associates the SIR (Signal to Interference power Ratio), which is an index of the channel state, with the data modulation method and channel coding rate, and the table is set according to the measured channel state. Referring to, AMC is realized by switching to the appropriate data modulation method and channel code rate.

FIG. 14 is a diagram illustrating an example of mapping from a channel coding block to a radio frame in the mapping unit 15 of FIG. Fig. 14 (a) shows the case of mapping from multiple channel coding blocks to radio frames, and Fig. 14 (b) shows the case of mapping from one channel coding block to multiple radio frames. Radio frames are divided into chunks that are transmission units. Chunks are frequency (OFDM: Orthogonal Frequency Division Multiplexing, etc.), code (CDMA: Code Division Multiple Access, etc.), and temporal (TDMA: Time Division Multiple Access etc.), spatial (MIMO: Multiple Input Multiple Output etc.) or a combination of these.

Next, FIGS. 15 to 18 are explanatory diagrams of retransmission control performed by the retransmission control unit 17 of the transmitter 10 and the retransmission control unit 23 of the receiver 20 in FIG.

[0051] FIG. 15 is a diagram showing hybrid ARQ processing, where the transmitter 10 side assigns CRC bits.

(Step S101) and error correction coding (Step S102), and after error correction decoding (Step S201) on the receiver 20 side, error detection using CRC bits is performed (Step S202). If there is an error, a retransmission request is sent to the transmitter 10, and if there is no error, transmission (reception) is completed (step S203). FIG. 16 is an explanatory diagram of a retransmission request signal error detection type hybrid ARQ based on the CRC code adopted in the embodiment of the present invention. As shown in FIG. The retransmission request signal is accompanied by other CRC symbols (CRC-bits) obtained by CRC coding of the ACKZNACK signal together with other control symbols (reception SIR required for adaptive modulation / demodulation / channel coding, etc.). The error of the signal itself is reduced. As a result, as shown in Fig. 16 (b), when packet #k is sent from the transmitter 10 side, an error is detected on the receiver 20 side and a NACK signal is sent to the transmitter 10 side. Even if it is received as an ACK signal by mistake, it is recognized as a NACK signal by the CRC code, and packet #k in which an error has occurred can be retransmitted appropriately. FIG. 17 is a diagram showing an example of a retransmission request signal error generated in the conventional general hybrid ARQ shown for comparison with the present embodiment. As shown in FIG. 17 (a), the original ACK signal is shown. If it is recognized as a NACK signal, it will retransmit the packet # k that would normally be received normally, reducing efficiency. Also, as shown in Fig. 17 (b), when the original NACK signal is recognized as an ACK signal, the packet is received normally and packet #k is treated as being normally received. Will be lost and the reception quality will be degraded. By the way, as shown in Fig.18, it was reached normally by the upper RLC layer, and it was dealt with by retransmitting the packet, but the retransmission of the MAC layer was performed in about 10 msec. On the other hand, retransmission of the RLC layer took about 100 msec, and the delay time was negligible. In this regard, in the present embodiment employing the retransmission request signal error detection hybrid ARQ based on the CRC code, the ACKZNACK signal can be accurately recognized, and high-quality transmission can be performed without using the RLC layer. it can

[0053] As described above, in the present embodiment, since the RLC layer header is eliminated, the overhead can be reduced and the efficiency can be improved. In addition, quality control performance can be improved by making the retransmission unit closer to the IP packet. Furthermore, by dividing and concatenating with several patterns according to the size of the IP packet, it is possible to appropriately cope with IP packets with a wide distribution size.

Next, when mapping without dividing the IP packet (FIG. 4 (a), P1 to P7 in FIG. 5, FIG. 6). The method for solving the problem of efficiency degradation due to retransmission in the above method is explained. This method can be applied to other patterns that are particularly effective when mapping IP packets without segmentation.

[0055] That is, in normal hybrid ARQ, when an error is detected, a NACK signal is returned to the transmission side to perform retransmission in units of channel coding blocks (ARQ blocks), and the channel is not divided into IP packets. When coding blocks are used, the efficiency decreases due to retransmission when IP packets are long. However, even if it contains errors, there are cases where the whole is wrong or only part is wrong, and it is not efficient to always retransmit the whole. Therefore, a method that performs the minimum necessary retransmission by adding the reliability indicating the degree of overall error or the reliability of each Z block for each bit to the NACK signal is adopted.

FIG. 19 is a diagram showing another configuration example of the communication system to which the IP packet mapping method of the present invention is applied. The configuration in FIG. 19 is almost the same as that shown in FIG. 10, and the receiver 20 detects the reliability for each bit from the demodulation process of the demodulator 21 and averages these as necessary for each block.ヽ generates a reliability indicating the degree of the overall error, adds a piece of information to the retransmission control unit 23, and returns a NACK signal. 10 is different in that the retransmission control unit 17 has a function of performing the minimum necessary retransmission based on the reliability added to the NACK signal.

FIG. 20 is a diagram showing an example of a NACK signal. (A) shows a case with information indicating reliability indicating the degree of error of the entire ARQ block together with identification information indicating that the signal is a NACK signal. b) shows a case with information indicating the reliability indicating the degree of error for each bit with an identification of being a NACK signal, _(c) together with identification information that is NACK signal error for each block The case where the information which shows the reliability which shows a grade is shown is shown. In the case of (b) and (c), the reliability information can be specified information of a location where the reliability is low and it can be estimated that an error has occurred.

[0058] Fig. 21 is a diagram showing an example of an ACKZNACK signal that also shows the overall reliability corresponding to Fig. 20 (a). An error-free ACK signal is regarded as one type of ACK (O), and an error is detected. NACK signal that indicates that there is a high degree of reliability, NACK (0) force and NACK (6) 7 A staged example is shown. Note that the number of NACK signals can be determined arbitrarily in consideration of control efficiency.

[0059] Also, in Fig. 20 (b) and (c), the reliability for each bit is shown. In (a), the amount of information to be fed back increases, so the reliability for each block is shown. Is more advantageous. FIG. 22 is a diagram showing an example of a block, in which one packet is divided into four blocks # 1 to # 4.

[0060] The retransmission control unit 17 of the transmitter 10 that has received the NACK signal accompanied by the reliability information described above retransmits a large number of bits when the reliability is low when the reliability information is the whole reliability information. When the reliability is high, only some bits are retransmitted. In addition, when combining packets with packets punctured with different patterns in the no and the hybrid ARQ, it is possible to cope with this by changing the size of the added redundancy. Figure 23 shows the type of hybrid ARQ processing. As shown in (a), if there is a demodulation error in packet pi, packet pi is discarded and packet p2 of the same content is received and retransmitted. As shown in (b) and (c), if there is a demodulation error in packet pi, the packet pi is retained without being discarded, and this packet pi is retransmitted. There is a type in which a packet p3 is generated by combining the received packet p2 and the packet p3 is demodulated. Here, (b) receives the same packet retransmission, and the received SIR is improved by packet combining. (C) receives a retransmission of a packet that has been punctured with a different pattern, and improves the coding gain by combining packets. When changing the degree of redundancy added above, it is of type (c).

In addition, the retransmission control unit 17 of the transmitter 10 that has received the NACK signal accompanied by reliability information generates an error with low reliability when the reliability information is for each bit or block. Retransmit only the bits or blocks that are determined to be active.

FIGS. 24 to 28 show a method of generating the reliability of the demodulation process power.

FIG. 24 is a diagram illustrating a configuration example of a decoder that performs iterative decoding, and y to y are demultiplexing values.

0 2

Received signal (decoder input signal), y is the only information bit of the signal sequence.

0

Y

1 is a signal whose signal sequence is recursively systematically convolutionally encoded, y

2 is a signal in which the signal sequence is interleaved and further recursively systematically convolutionally coded. And trust The signal y and the signal y are decoded by the soft input / soft output decoder DEC1, and the output signal and signal are decoded.

0 1

A soft input / soft output decoder where y is interleaved by interlino I and signal y

0 2

Decoded by DEC2 and the output signal deinterleaved by dintariba DI are added to the input signal of soft input / soft output decoder DEC1. The output signals of the soft input / soft output decoders DEC1 and DEC2 are hard-decided by the discriminators Dl and D2 to become output signals.

FIG. 25 is a diagram showing an example of detecting the reliability using the soft decision output of the decoder, and adopts the soft decision output (likelihood) for each bit included in the bucket as an index of reliability. be able to. That is, the iterative decoder as shown in FIG. 24 has a configuration in which a plurality of soft input / soft output decoders DEC1 and DEC2 are connected, and reproduces a signal sequence by making a hard decision on the soft output at the final stage. Yes, close to the threshold level (0 level) for hard decision, el and e2 with small absolute values have high probability of error and low reliability.

FIG. 26 is a diagram showing an example of detecting reliability using the number of code inversions in the iterative decoding process. The number of times that the soft decision output for each bit included in the packet inverts the code in the iterative process is shown. It can be adopted as an index of reliability. In other words, el and e2, which are sign-inverted between the number of repetitions i and i + 1, have low reliability because the decoding result is not stable and the probability of error is high.

[0066] FIG. 27 is a diagram illustrating an example of calculation processing when reliability is detected using cross-entropy in iterative decoding, and can be calculated using the signal values shown in FIG. . In this case, the larger the cross-entropy T, the more stable the decoding result, and

CE

Since the rate is high, the reliability is low.

[0067] FIG. 28 is a diagram illustrating an example of detecting reliability using reception SINR. Since the reception SINR for each bit included in the packet is smaller, the probability of error is higher, and thus the reliability is higher. Low!

[0068] Also, the lowest log likelihood ratio (LLR) in a packet can be used as an index of reliability.

[0069] The present invention has been described above with reference to preferred embodiments of the present invention. Although the invention has been described herein with reference to specific embodiments, the broad scope of the invention as defined in the claims. It is clear that various modifications and changes can be made to these examples without departing from the spirit and scope. In other words, the present invention should not be construed as being limited by the details of the specific examples and the accompanying drawings.

This international application has priority based on Japanese Patent Application No. 2005-106906 filed on April 1, 2005 and Japanese Patent Application No. 2005-174402 filed on June 14, 2005. All of which are incorporated herein by reference.

Claims

The scope of the claims

[1] If the IP packet is smaller than the predetermined size, leave it as it is, and if it exceeds the predetermined size, generate a MAC-PDU by dividing it,

A process of adding a header including IP packet division information and a CRC code to the MAC-PDU;

A MAC coding PDU with the header and CRC code attached thereto and a channel code to generate a channel coding block;

And a step of mapping the channel coding block to a radio frame.

[2] generating one or more fixed-size MAC-PDUs from the IP packet, performing a header including the IP packet division information on the MAC-PDU, and adding a CRC code to the MAC-PDU;

And a step of mapping the channel coding block to a radio frame.

[3] Creating a variable-size MAC PDU from an IP packet as is,

And a step of mapping the channel coding block to a radio frame.

[4] The IP packet mapping method according to claim 1,

The IP packet mapping method, wherein the predetermined size is adaptively set according to reception quality information and Z or a frequency of retransmission request signals.

[5] The IP packet mapping method according to claim 1,

When the IP packet exceeds the specified size, the IP packet is divided from the beginning. An IP packet mapping method, wherein the packet is divided at the predetermined size.

[6] In the IP packet mapping method according to claim 1,

An IP packet mapping method characterized in that, when the IP packet exceeds a predetermined size, the IP packet is equally divided into a predetermined number so that the IP packet is equal to or smaller than the predetermined size.

[7] The IP packet mapping method according to claim 1,

Receiving side power receiving a reliability indicating the degree of error of the channel coding block;

According to the above reliability, the method includes: a step of retransmitting a large number of bits when the reliability is low, and a step of retransmitting only a part of the bits when the reliability is high. .

[8] According to the IP packet mapping method of claim 7,

The IP packet mapping method, wherein the reliability is received by a NACK signal divided into levels according to the reliability.

[9] According to the IP packet mapping method of claim 1,

Receiving a reliability indicating the degree of error for each bit or block of the channel coding block from the receiving side;

And a step of retransmitting only a bit or a block determined to have an error due to low reliability according to the reliability.

[10] According to the IP packet mapping method of claim 1,

If the size of the IP packet as it is or divided is less than a predetermined value or if there is a margin in size, connection with the following IP packet, rate matching by repeated encoding, or dummy bit insertion is performed. An IP packet mapping method characterized by the above.

[11] In the IP packet mapping method according to claim 1,

The IP packet mapping method, wherein the retransmission request signal from the receiving side based on the CRC code is accompanied by another CRC code obtained by CRC coding the ACKZNACK signal together with other control symbols. Split the IP packet and concatenate with Z or the following IP packet to generate a MAC-PDU

The MAC-PDU includes a packet code header for performing a header and a CRC code with IP packet segmentation information,

A channel code key part for generating a channel coding block by channel-coding the MAC-PDU with the header and CRC code attached thereto;

A transmitter comprising: a mapping unit that maps the channel coding block to a radio frame.