CN104158628B - Relay transfer method based on unique decodable code - Google Patents

Abstract

Based on the unique decodable relay forwarding method, the present invention relates to the uniquely decodable based relay forwarding method. The present invention aims to solve the problem that the number of time slots to be transmitted increases and the throughput of the system decreases due to the high bit error rate of the Rayleigh channel and the relay system of M (M>3) nodes in the existing physical layer network coding. For a M-node relay system, including M-1 user nodes and a relay node, any user node uses two time slots to obtain information sent by other arbitrary user nodes through the relay node; at the first In the second time slot, all M-1 user nodes will send the uniquely decodable and encoded information to the relay node; in the second time slot, the relay node will judge and forward the judgment result and broadcast Each user node matches the received broadcast information with a unique decodable code. The invention is applied in the communication field.

Description

Relay and forwarding method based on unique decodability

technical field

The invention relates to a unique decodable-based relay forwarding method.

Background technique

With the development of science and technology, Network Coding (NC) has gradually become a popular research field in the communication field. Network coding can be used in wireless network systems and wired network systems, and has a great impact on information theory, wireless communication, network architecture, etc. Physical-layer network coding (PNC) has gradually become an important research hotspot in the field of network coding. PNC improves the throughput of the system by effectively utilizing the properties of relay nodes and reducing the number of time slots that need to be transmitted between nodes. Not only that, PNC can be regarded as a cooperative communication.

The research on PNC mainly focuses on the following directions. First of all, research on PNC coding methods, including convolutional codes, low density parity check codes (Low Density Parity Check Code, LDPC), LDPC accumulation codes and so on. Secondly, research on PNC modulation methods includes different modulation methods, including frequency-shift keying modulation (Frequency-shiftkeying, FSK), quadrature amplitude modulation (Quadrature Amplitude Modulation, QAM) and so on. In addition, there are studies on synchronization issues, channel capacity, etc. in PNC.

Most of the existing research on PNC is under Gaussian channel (additive white Gaussian noise, AWGN), and PNC has excellent performance under Gaussian channel. However, in the Rayleigh channel, PNC will be difficult to perform, and the existing solution is to use a multiple-input multiple-output (MIMO) system. But this will waste the degree of freedom gain brought by MIMO technology. Moreover, in an M (M>3) node relay system, including M-1 user nodes and 1 relay node, it will be difficult for the PNC to implement exclusive OR (XOR) operations.

Contents of the invention

The present invention aims to solve the problem that the number of time slots to be transmitted increases and the throughput of the system decreases when the bit error rate of the existing physical layer network coding is high in the Rayleigh channel and in the relay system of M (M>3) nodes. Instead, a unique decodable-based relay forwarding method is provided.

The relay forwarding method based on uniquely decodable is implemented in the following steps:

For an M-node relay system, including M-1 user nodes and a relay node, any user node uses two time slots to obtain information sent by any other user node through the relay node;

Among them, in the first time slot, all M-1 user nodes will send the uniquely decodable and encoded information to the relay node;

Then in the second time slot, the relay node sends the judged result in the form of broadcast after judgment and forwarding, and each user node matches the received broadcast information with a unique decodable code, thereby Obtaining the information sent by other user nodes completes the unique decodable-based relay forwarding method.

Invention effect:

The present invention aims at the shortcomings of PNC in the multi-node relay system, starts from the perspective of coding, applies the unique decodable code to the physical layer, and maps different user terminals through the data information of the relay nodes, thereby improving the system efficiency. throughput. Not only that. The relay forwarding strategy based on the unique decodability proposed by the present invention has good bit error rate performance even in the Rayleigh channel.

Under the Gaussian channel, when the channel capacity of the present invention is M=3, when the signal-to-noise ratio (SNR>8dB), there will be a qualitative improvement relative to the PNC, and when M=3, the upper limit of the channel capacity of the PNC will be broken. (0.5), and using the unique decodable with better performance can even make the upper bound of the channel capacity close to 0.75. When M>3, using the only decodable relay and forwarding system can greatly reduce the transmission time slot, and it can be completed by 2 time slots, which greatly improves the throughput of the system.

Based on the uniquely decodable relay forwarding strategy proposed by the present invention, in the case of M=3, when time slot 1 becomes a Rayleigh channel, it still has acceptable symbol error performance and has better channel capacity. It can still break through the theoretical upper limit of PNC channel capacity of 0.5.

Description of drawings

FIG. 1 is a block diagram of a relay system of an M node in Embodiment 1;

Fig. 2 is a system block diagram of time slot 1 in the second embodiment;

Fig. 3 is a system block diagram of time slot 2 in the third embodiment;

Fig. 4 is the flow chart of the time slot 1 system in the second embodiment;

Fig. 5 is a flow chart of the time slot 2 system in the third embodiment;

Fig. 6 is a bit error rate simulation diagram in the simulation experiment;

Fig. 7 is a throughput simulation diagram in the simulation experiment.

detailed description

Specific implementation mode 1: The uniquely decodable-based relay forwarding method of this implementation mode is implemented in the following steps:

For an M-node relay system, including M-1 user nodes and a relay node, any user node uses two time slots to obtain information sent by any other user node through the relay node;

Among them, in the first time slot, all M-1 user nodes will send the uniquely decodable and encoded information to the relay node;

Then in the second time slot, the relay node sends the judged result in the form of broadcast after judgment and forwarding, and each user node matches the received broadcast information with a unique decodable code, thereby Obtaining the information sent by other user nodes completes the unique decodable-based relay forwarding method.

Specific implementation mode two: the difference between this implementation mode and specific implementation mode one is: the execution steps of the first time slot are as follows:

Step 1, all M-1 user nodes pass the effective information data of the user nodes through a unique decodable encoder to obtain encoded codewords d ₁ , d ₂ ,...,d _M-1 ; where, the Said each codeword is an n-dimensional vector;

Step 2: Modulate the M-1 codewords d ₁ , d ₂ ,...,d _M-1 obtained in step 1 into s ₁ , s ₂ ,...,s _M-1 , where s _i represents the modulated symbol sent by the i-th user in the first time slot, where i=1,2,...,M-1;

Step 3, pass s ₁ , s ₂ ,...,s _M-1 in step 2 through channels h ₁ , h ₂ ,..., h _M-1 respectively, and receive signal r _R at the relay node;

in,

Wherein, the w _R represents the Gaussian white noise at the relay node in the first time slot, and w _R is an n-dimensional vector;

For the channels h ₁ , h ₂ ,...,h _M-1 , if h ₁ =h ₂ =...=h _M-1 =1, then in time slot 1, the channels of each user and relay node is the Gaussian channel;

If the channels h ₁ , h ₂ ,...h _M-1 all obey the independent and identically distributed Rayleigh distribution, then the channel of each user and relay node is a Rayleigh channel;

Step 4, judge r _R as a judged symbol through the judger of the relay node

Step five, calculate the set Among them, h _i represents the channel parameter from the i-th user to the relay node in the first time slot, s _i is an n-dimensional vector, and i represents the i-th user;

Step six, judge Whether it belongs to R _Relay , if yes, the time slot is completed, if not, transfer to step 3, pass s ₁ , s ₂ ,...,s _M-1 through channels h ₁ , h ₂ ,... ,h _M-1 is resent, and the signal r _R is received at the relay node.

Other steps and parameters are the same as those in Embodiment 1.

Specific implementation mode three: the difference between this implementation mode and specific implementation modes one or two is that: the execution steps of the second time slot are as follows:

Step one, will The relay node broadcasts through channels h ₁ ′, h ₂ ′,…h′ _M-1 ; wherein, the channels h ₁ ′, h ₂ ′,…h′ _M-1 , if h ₁ '=h ₂ '=...=h' _M-1 =1, then in the second time slot, the broadcast channel between the relay node and each user is the Rayleigh channel;

If the channels h ₁ ′, h ₂ ′,...h′ _M-1 all obey the independent and identically distributed Rayleigh distribution, then in the second time slot, the broadcast channel of the relay node and each user is a Rayleigh channel;

Step 2: Among all M-1 users, judge the received broadcast information r ₁ , r ₂ ,...,r _M-1 as

Among them, r _i represents the signal received by the i-th user in the second time slot, then r _i (1≤i≤M-1) is

Among them, w _i (1≤i≤M-1) is independent and identically distributed Gaussian white noise, w _i , r _i are both n-dimensional vectors;

Indicates the result after the signal r _i received by the i-th user is judged in the second time slot;

Step three, calculate h _i 'indicates that the second time slot is from the relay node to the i-th user channel;

Step four, judge Are they all in the R _user ? If yes, go to step five and the second time slot ends; otherwise, go to step one and set Resend in the form of broadcast through the relay node;

Step 5, on the M-1 user nodes respectively set Through the unique decodable encoder mapping, decoded into Time slot two ends.

Other steps and parameters are the same as those in Embodiment 1 or Embodiment 2.

Specific implementation mode four: the difference between this implementation mode and one of specific implementation modes one to three is that the unique decodable code is specifically:

In the M-node relay system, it is assumed that the unique decodable word set for the i-th user (1≤i≤M-1) is C _i , and it is assumed that in the code word set of each user, there is |C _i | codewords, each codeword length is n, from the definition of unique decodability, in the i-th user, any two different codewords u _i and u _i ′, need to satisfy

Wherein, both ends of formula (3) are an n-dimensional vector, and the vector is M-ary;

Let the set E be the set of bitwise sums of all possible codewords of length n for M-1 user nodes, and the set E is expressed as Where i=1,2,...,M-1, according to the definition of uniquely decodable, we know that there are elements, each of which is an n-dimensional M-ary vector;

The decoding process is as follows, due to the uniquely decodable nature of M-1 users, knowing the sum of any M-1 codewords can be uniquely mapped to these M-1 codewords, that is, the only decodable The decoding method is to restore M-1 codewords through mapping according to the sum vector of the codewords.

Other steps and parameters are the same as those in Embodiments 1 to 3.

Specific embodiment five: this embodiment is different from one of specific embodiments one to four in that: if the modulation method in the execution step 2 of the first time slot is BPSK, then

s _i =2·d _i -I,(i=1,2,...,M-1) (4)

Wherein, I is an n-dimensional all-one vector, that is, I=(1,1,...,1).

Other steps and parameters are the same as in one of the specific embodiments 1 to 4.

Embodiment 6: This embodiment is different from Embodiment 1 to Embodiment 5 in that: the judgment is a hard judgment or a soft judgment. Other steps and parameters are the same as one of the specific embodiments 1 to 5.

Simulation:

(1) Gaussian channel

Under the Gaussian channel, the simulation parameters are shown in Table 1.

Table 1 Gaussian channel simulation parameters

The R _Relay at this time is {(-2,-2),(0,0),(-2,0),(0,2),(0,-2),(2,0)}. In the same way, the R _user at this time is {(-2,-2),(0,0),(-2,0),(0,2),(0,-2),(2,0 )}. The results of bit error rate simulation and channel capacity simulation are shown in Fig. 6 and Fig. 7 respectively.

As shown in Fig. 6 and Fig. 7, under the Gaussian channel, though, the symbol error performance of the uniquely decodable relay forwarding strategy proposed by the present invention is slightly worse than that of PNC. However, the channel capacity of the present invention, when the signal-to-noise ratio is large (SNR>8dB), will have a qualitative improvement relative to the PNC, and will break through the upper limit (0.5) of the channel capacity of the PNC when M=3. Using the unique decodable with better performance can even make the upper bound of the channel capacity close to 0.75.

(2) Rayleigh channel

Under the Rayleigh channel, the simulation parameters are shown in Table 2.

Table 2 Rayleigh channel simulation parameters

For the convenience of expression, let x=h ₁ +h ₂ , y=h ₁ -h ₂ . The R _Relay at this time is {(-x,-x),(y,y),(-x,-y),(y,x),(-y,-x),(x,y)}. In the same way, the R _user at this time is {(-x,-x),(y,y),(-x,-y),(y,x),(-y,-x),(x ,y)}. The results of bit error rate simulation and channel capacity simulation are shown in Table 3 respectively.

Table 3 The performance of this simulation experiment under the Rayleigh channel

As can be seen from Table 3, based on the uniquely decodable relay forwarding strategy proposed by the present invention, in the case of M=3, when time slot 1 becomes a Rayleigh channel, there is still acceptable symbol error performance, and With better channel capacity, it can still break through the theoretical upper bound of PNC channel capacity of 0.5.

To sum up, the uniquely decodable-based relay forwarding strategy proposed by the present invention can be applied in the M-node relay system, greatly improving the throughput of the system, and ensuring that the bit error rate is within an acceptable range.

Claims (5)

1. Based on the uniquely decodable relay forwarding method, it is characterized in that based on the uniquely decodable relay forwarding method is realized in the following steps: For an M-node relay system, including M-1 user nodes and a relay node, any user node uses two time slots to obtain information sent by any other user node through the relay node; Among them, in the first time slot, all M-1 user nodes will send the uniquely decodable and encoded information to the relay node; Then in the second time slot, the relay node sends the judged result in the form of broadcast after judgment and forwarding, and each user node matches the received broadcast information with a unique decodable code, thereby Obtain the information sent by other user nodes, that is, complete the relay forwarding method based on unique decodability; The execution steps of the first time slot are as follows: Step 1, all M-1 user nodes pass the effective information data of the user nodes through a unique decodable encoder to obtain encoded codewords d ₁ , d ₂ ,...,d _M-1 ; where, each A codeword is an n-dimensional vector; Step 2: Modulate the M-1 codewords d ₁ , d ₂ ,...,d _M-1 obtained in step 1 into s ₁ , s ₂ ,...,s _M-1 , where s _i represents the modulated symbol sent by the i-th user in the first time slot, where i=1,2,...,M-1; Step 3, pass s ₁ , s ₂ ,...,s _M-1 in step 2 through channels h ₁ , h ₂ ,..., h _M-1 respectively, and receive signal r _R at the relay node; in, Wherein, the w _R represents the Gaussian white noise at the relay node in the first time slot, and w _R is an n-dimensional vector; For the channels h ₁ , h ₂ ,...,h _M-1 , if h ₁ =h ₂ =...=h _M-1 =1, then in time slot 1, the channels of each user and relay node is the Gaussian channel; If the channels h ₁ , h ₂ ,...h _M-1 all obey the independent and identically distributed Rayleigh distribution, then the channel of each user and relay node is a Rayleigh channel; Step 4, judge r _R as a judged symbol through the judger of the relay node Step five, calculate the set Among them, h _i represents the channel parameter from the i-th user to the relay node in the first time slot, s _i is an n-dimensional vector, and i represents the i-th user; Step six, judge Whether it belongs to R _Relay , if yes, the time slot is completed, if not, transfer to step 3, pass s ₁ , s ₂ ,...,s _M-1 through channels h ₁ , h ₂ ,... ,h _M-1 is resent, and the signal r _R is received at the relay node. 2. The uniquely decodable relay forwarding method according to claim 1, characterized in that the execution steps of the second time slot are as follows: Step one, will The relay node broadcasts through channels h ₁ ′, h ₂ ′,…h′ _M-1 ; wherein, the channels h ₁ ′, h ₂ ′,…h′ _M-1 , if h ₁ '=h ₂ '=...=h' _M-1 =1, then in the second time slot, the broadcast channel between the relay node and each user is the Rayleigh channel; If the channels h ₁ ′, h ₂ ′,...h′ _M-1 all obey the independent and identically distributed Rayleigh distribution, then in the second time slot, the broadcast channel of the relay node and each user is a Rayleigh channel; Step 2: Among all M-1 users, judge the received broadcast information r ₁ , r ₂ ,...,r _M-1 as Among them, r _i represents the signal received by the i-th user in the second time slot, then r _i (1≤i≤M-1) is Among them, w _i (1≤i≤M-1) is independent and identically distributed Gaussian white noise, w _i , r _i are both n-dimensional vectors; Indicates the result after the signal r _i received by the i-th user is judged in the second time slot; Step three, calculate h _i 'indicates that the second time slot is from the relay node to the i-th user channel; Step four, judge Are they all in the R _user ? If yes, go to step five and the second time slot ends; otherwise, go to step one and set Resend in the form of broadcast through the relay node; Step 5, on the M-1 user nodes respectively set Through the unique decodable encoder mapping, decoded into Time slot two ends. 3. The uniquely decodable-based relay forwarding method according to claim 2, wherein the uniquely decodable code is specifically: In the M-node relay system, it is assumed that the unique decodable word set for the i-th user (1≤i≤M-1) is C _i , and it is assumed that in the code word set of each user, there is |C _i | codewords, each codeword length is n, from the definition of unique decodability, in the i-th user, any two different codewords u _i and u _i ′, need to satisfy Among them, both ends of formula 3 are an n-dimensional vector, and the vector is M-ary; Let the set E be the set of bitwise sums of all possible codewords of length n for M-1 user nodes, and the set E is expressed as Where i=1,2,...,M-1, according to the definition of uniquely decodable, we know that there are elements, each of which is an n-dimensional M-ary vector; The decoding process is as follows, due to the uniquely decodable nature of M-1 users, knowing the sum of any M-1 codewords can be uniquely mapped to these M-1 codewords, that is, the only decodable The decoding method is to restore M-1 codewords through mapping according to the sum vector of the codewords. 4. The method of relaying and forwarding based on unique decodability according to claim 2, wherein if the modulation method in the execution step 2 of the first time slot is BPSK, then s _i =2·d _i -I,(i=1,2,...,M-1) Formula 4 Wherein, I is an n-dimensional all-one vector, that is, I=(1,1,...,1). 5. The uniquely decodable-based relay forwarding method according to claim 3, characterized in that the decision is a hard decision or a soft decision.