CN104158628A - Relaying method based on uniquely decodable code - Google Patents

Abstract

The present invention relates to a relaying method based on uniquely decodable code. The present invention aims to solve the problem of existing physical-layer network coding about high error rate in Rayleigh channel and increased amount of time slots to be transmitted and decreased throughput of system in a system with M (M>3) nodes. For a relay system with M nodes comprising M-1 user nodes and one relay node, any one user node uses two time slots to obtain messages sent from any other user nodes through the relay node; in the first time slot, all of the M-1 user nodes send the messages coded by a uniquely decodable code to the relay node; and in the second time slot, the relay node sends decision results after deciding and forwarding in the form of broadcasting, and each user node matches the received broadcasting messages with the uniquely decodable code. The present invention is applied in communication field.

Description

Relay transfer method based on unique decodable code

Technical field

The present invention relates to the relay transfer method based on unique decodable code.

Background technology

Along with scientific and technical development, in the communications field, network code (Network Coding, NC) becomes a popular research field gradually.Network code can be in Radio Network System and wired network system, and gives information theory, radio communication, and the aspects such as the network architecture are with very large impact.And physical-layer network coding (physical-layer network coding, PNC) has become an important research focus in network code field gradually.PNC, by effectively utilizing the character of via node, by reducing the number of timeslots that needs transmission between node, thereby has improved the throughput of system, and moreover, PNC can be regarded as a kind of collaboration communication.

About the research of PNC, mainly concentrate on following direction.First, the research about in the coded system of PNC, comprises convolution code, low density parity check code (Low Density Parity Check Code, LDPC), LDPC accumulation code etc.Secondly, about the research of PNC modulator approach, comprise different modulation systems, comprise shift keying modulation (Frequency-shift keying, FSK), quadrature amplitude modulation (Quadrature Amplitude Modulation, QAM) etc.In addition, also relevant for stationary problem in PNC, the research of the problems such as channel capacity.

The existing research great majority about PNC are all under Gaussian channel (additive white Gaussian noise, AWGN), and PNC has outstanding performance under Gaussian channel.And in Rayleigh channel, PNC will be difficult to carry out, and existing solution is to utilize multiaerial system (Multiple-Input Multiple-Output, MIMO).But can slattern like this degree of freedom gain that MIMO technology is brought.Moreover, in M (M>3) node relay system, comprise M-1 user node and 1 via node, now PNC will be difficult to realize XOR (XOR) computing.

Summary of the invention

The present invention is that will to solve existing physical-layer network coding high and in the relay system of M (M>3) node in the Rayleigh channel error rate, need the number of timeslots of transmission to increase, the problem that the throughput of system reduces, and the relay transfer method based on unique decodable code is provided.

Relay transfer method based on unique decodable code is realized according to the following steps:

For the relay system of a M node, comprise M-1 user node and a via node, any one user node utilizes two time slots to obtain by via node the information that other any user nodes send;

Wherein, in first time slot, all M-1 user node, by the information of carrying out after unique decodable code coding, sends to via node;

Then in second time slot, via node is through judgement and forwarding, form the result after judgement with broadcast sends, the broadcast message that each user node receives by handle is mated with the decoding of unique decodable code, thereby obtain the information that all the other user nodes send, completed the relay transfer method based on unique decodable code.

Invention effect:

The present invention is directed to the shortcoming of PNC in multinode relay system, from the angle of coding, start with, unique decodable code is applied in to physical layer, the data message by via node, shines upon different user terminals, thereby has improved the throughput of system.Moreover.The relay forwarding strategy based on unique decodable code that the present invention proposes, even in Rayleigh channel, also has good bit error rate performance.

Under Gaussian channel, channel capacity of the present invention, during with M=3, when signal to noise ratio is larger (SNR>8dB), with respect to PNC, will there is the lifting of matter, in the time of will breaking through M=3, the channel capacity upper limit (0.5) of PNC, and adopt the better unique decodable code of performance, even can allow the upper bound of channel capacity approach 0.75.And when M>3, utilize the relay forwarding system of unique decodable code, and can significantly reduce transmission time slot, and can complete by 2 time slots, greatly promoted the throughput of system.

The relay forwarding strategy based on unique decodable code that the present invention proposes, the in the situation that of M=3, during 1 one-tenth Rayleigh channel of time slot, still have acceptable mistake symbol performance, and there is good channel capacity, now still can break through the theoretical upper bound 0.5 of PNC channel capacity.

Accompanying drawing explanation

Fig. 1 is the relay system block diagram of the M node in embodiment one;

Fig. 2 is time slot 1 system block diagram in embodiment two;

Fig. 3 is time slot 2 system block diagrams in embodiment three;

Fig. 4 is time slot 1 system flow chart in embodiment two;

Fig. 5 is time slot 2 system flow charts in embodiment three;

Fig. 6 is BER Simulation figure in emulation experiment;

Fig. 7 is throughput analogous diagram in emulation experiment.

Embodiment

Embodiment one: the relay transfer method based on unique decodable code of present embodiment is realized according to the following steps:

For the relay system of a M node, comprise M-1 user node and a via node, any one user node utilizes two time slots to obtain by via node the information that other any user nodes send;

Wherein, in first time slot, all M-1 user node, by the information of carrying out after unique decodable code coding, sends to via node;

Then in second time slot, via node is through judgement and forwarding, form the result after judgement with broadcast sends, the broadcast message that each user node receives by handle is mated with the decoding of unique decodable code, thereby obtain the information that all the other user nodes send, completed the relay transfer method based on unique decodable code.

Embodiment two: present embodiment is different from embodiment one: the execution step of described first time slot is as follows:

Step 1, all M-1 user node, passes through unique decodable code encoder, the code word d after being encoded by the effective information data of user node ₁, d ₂..., d _m-1; Wherein, described each code word is a n-dimensional vector;

Step 2, by the M-1 obtaining in step 1 code word d ₁, d ₂..., d _m-1, be modulated into respectively s ₁, s ₂..., s _m-1, s wherein _irepresent in first time slot, the symbol after the modulation of i user's transmission, wherein, and i=1,2 ..., M-1;

Step 3, by the s in step 2 ₁, s ₂..., s _m-1pass through respectively channel h ₁, h ₂..., h _m-1, at via node, receive signal r _r;

Wherein, $r_R=\underset{i=1}{\overset{M-1}{\mathrm{\Σ}}}{s}_i\·h_i+w_R---\left(1\right)$

Wherein, described w _rbe illustrated in first time slot the white Gaussian noise at via node place, and w _rfor n-dimensional vector;

Described channel h ₁, h ₂..., h _m-1if, h ₁=h ₂=...=h _m-1=1, in time slot 1, the channel of each user and via node is exactly Gaussian channel so;

If channel h ₁, h ₂... h _m-1be all to obey independent identically distributed rayleigh distributed, the channel of each user and via node is exactly Rayleigh channel so;

Step 4, by r _rdecision device judgement by via node is the symbol after adjudicating

Step 5, set of computations wherein, h _irepresent first time slot by i user the channel parameter to via node, s _ibe a n-dimensional vector, i represents i user;

Step 6, judgement whether belong to R _relay, if so, this time slot completes, and if not, is adjusted to step 3, by s ₁, s ₂..., s _m-1pass through respectively channel h ₁, h ₂..., h _m-1resend, at via node, receive signal r _r.

Other step and parameter are identical with embodiment one.

Embodiment three: present embodiment is different from embodiment one or two: the execution step of described second time slot is as follows:

Step 1, will by via node, with the form of broadcast, pass through channel h ' ₁, h ' ₂... h ' _m-1send; Wherein, described channel h ' ₁, h ' ₂... h ' _m-1if, h ' ₁=h ' ₂=...=h ' _m-1=1, in the 2nd time slot, via node and each user's broadcast channel is exactly Rayleigh channel so;

If channel h ' ₁, h ' ₂... h ' _m-1be all to obey independent identically distributed rayleigh distributed, in the 2nd time slot, via node and each user's broadcast channel is exactly Rayleigh channel so;

Step 2, in all M-1 user, the broadcast message r receiving ₁, r ₂..., r _m-1adjudicate into wherein, r _irepresentative in second time slot, i the signal that user receives, r _i(1≤i≤M-1) is

$r_i=h_i^{\′}\·{\hat{s}}_R+w_i---\left(2\right)$

Wherein, w _i(1≤i≤M-1) is independent identically distributed white Gaussian noise, w _i, r _ibe n-dimensional vector;

be illustrated in second time slot i the signal r that user receives _iresult after being judged;

Step 3, calculates h ' _irepresent second time slot by via node to i subscriber channel;

Step 4, judgement whether all at R _userin, if so, carrying out step 5, second time slot finishes; Otherwise, be adjusted to step 1, will form by via node with broadcast resends;

Step 5, on M-1 user node respectively by unique decodable code encoder, shine upon, be decoded into time slot two finishes.

Other step and parameter are identical with embodiment one or two.

Embodiment four: present embodiment is different from one of embodiment one to three: described unique decodable code coding is specially:

In the relay system of M node, suppose that the unique decodable code word set for i user (1≤i≤M-1) is C _i, and suppose, in each user's codeword set, to have | C _i| individual code word, each code word size is n, from the definition of unique decodable code, in i user, appoints and gets two different code word u _iand u ' _i, need to meet

${\mathrm{\Σ}}_{i=1}^{M-1}{u}_i\≠{\mathrm{\Σ}}_{i=1}^{M-1}{u_i}^{\′}---\left(3\right)$

Wherein, formula (1) two ends are all n-dimensional vectors, and vector is M system;

Order set E, for the step-by-step of all possible length of M-1 the user node code word that is n do and set, gather E and be expressed as i=1 wherein, 2 ..., M-1, according to the definition of unique decodable code, has in known set E individual element, the M system vector that each element is n dimension;

As follows in decode procedure, due to the character of M-1 user's unique decodable code, the known code word of M-1 arbitrarily and, just can be mapped to uniquely this M-1 code word, the decoded mode that is unique decodable code is according to code word and vectorial, by mapping, restores M-1 code word.

Other step and parameter are identical with one of embodiment one to three.

Embodiment five: present embodiment is different from one of embodiment one to four: modulation system in the execution step two of described first time slot is if BPSK, so

s _i＝2·d _i-I,(i＝1,2,...,M-1) (4)

Wherein, I be n tie up complete 1 vector be I=(1,1 ..., 1).

Other step and parameter are identical with one of embodiment one to four.

Embodiment six: present embodiment is different from one of embodiment one to five: described judgement is hard decision or soft-decision.Other step and parameter are identical with one of embodiment one to five.

Emulation experiment:

(1) Gaussian channel

Under Gaussian channel, simulation parameter is as shown in table 1.

Table 1 Gaussian channel simulation parameter

R now _relayfor { (2 ,-2), (0,0), (2,0), (0,2), (0 ,-2), (2,0) }.In like manner can ask R now _userfor { (2 ,-2), (0,0), (2,0), (0,2), (0 ,-2), (2,0) }.The result of the emulation of BER Simulation and channel capacity respectively as shown in Figure 6 and Figure 7.

Shown in Fig. 6 and Fig. 7, under Gaussian channel, although the mistake symbol slightly inferior properties of the relay forwarding strategy based on unique decodable code that the present invention proposes is in PNC.But channel capacity of the present invention, when signal to noise ratio is larger (SNR>8dB), will have the lifting of matter with respect to PNC, in the time of will breaking through M=3, the channel capacity upper limit (0.5) of PNC.And adopt the better unique decodable code of performance, even can allow the upper bound of channel capacity approach 0.75.

(2) Rayleigh channel

Under Rayleigh channel, simulation parameter is as shown in table 2.

Table 2 Rayleigh channel simulation parameter

For convenient, represent, might as well make x=h ₁+ h ₂, y=h ₁-h ₂.R now _relayfor { (x ,-x), (y, y), (x ,-y), (y, x), (y ,-x), (x, y) }.In like manner can ask R now _userfor { (x ,-x), (y, y), (x ,-y), (y, x), (y ,-x), (x, y) }.The result of the emulation of BER Simulation and channel capacity is as shown in table 3 respectively.

Performance under this emulation experiment of table 3 Rayleigh channel

As can be seen from Table 3, the relay forwarding strategy based on unique decodable code that the present invention proposes, the in the situation that of M=3, during 1 one-tenth Rayleigh channel of time slot, still have acceptable mistake symbol performance, and there is good channel capacity, now still can break through the theoretical upper bound 0.5 of PNC channel capacity.

To sum up, the relay forwarding strategy based on unique decodable code that the present invention proposes can be applied in the relay system of M node, and the significantly throughput of elevator system, and the assurance error rate is within acceptable scope.

Claims (6)

1. the relay transfer method based on unique decodable code, is characterized in that the relay transfer method based on unique decodable code is realized according to the following steps:

For the relay system of a M node, comprise M-1 user node and a via node, any one user node utilizes two time slots to obtain by via node the information that other any user nodes send;

Wherein, in first time slot, all M-1 user node, by the information of carrying out after unique decodable code coding, sends to via node;

Then in second time slot, via node is through judgement and forwarding, form the result after judgement with broadcast sends, the broadcast message that each user node receives by handle is mated with the decoding of unique decodable code, thereby obtain the information that all the other user nodes send, completed the relay transfer method based on unique decodable code.

2. the relay transfer method based on unique decodable code according to claim 1, is characterized in that the execution step of described first time slot is as follows:

Step 1, all M-1 user node, passes through unique decodable code encoder, the code word d after being encoded by the effective information data of user node ₁, d ₂..., d _m-1; Wherein, described each code word is a n-dimensional vector;

Step 2, by the M-1 obtaining in step 1 code word d ₁, d ₂..., d _m-1, be modulated into respectively s ₁, s ₂..., s _m-1, s wherein _irepresent in first time slot, the symbol after the modulation of i user's transmission, wherein, and i=1,2 ..., M-1;

Step 3, by the s in step 2 ₁, s ₂..., s _m-1pass through respectively channel h ₁, h ₂..., h _m-1, at via node, receive signal r _r;

Wherein, $r_R=\underset{i=1}{\overset{M-1}{\mathrm{\Σ}}}{s}_i\·h_i+w_R---\left(1\right)$

Wherein, described w _rbe illustrated in first time slot the white Gaussian noise at via node place, and w _rfor n-dimensional vector;

Described channel h ₁, h ₂..., h _m-1if, h ₁=h ₂=...=h _m-1=1, in time slot 1, the channel of each user and via node is exactly Gaussian channel so;

If channel h ₁, h ₂... h _m-1be all to obey independent identically distributed rayleigh distributed, the channel of each user and via node is exactly Rayleigh channel so;

Step 4, by r _rdecision device judgement by via node is the symbol after adjudicating

Step 5, set of computations wherein, h _irepresent first time slot by i user the channel parameter to via node, s _ibe a n-dimensional vector, i represents i user;

Step 6, judgement whether belong to R _relay, if so, this time slot completes, and if not, is adjusted to step 3, by s ₁, s ₂..., s _m-1pass through respectively channel h ₁, h ₂..., h _m-1resend, at via node, receive signal r _r.

3. the relay transfer method based on unique decodable code according to claim 1 and 2, is characterized in that the execution step of described second time slot is as follows:

Step 1, will by via node, with the form of broadcast, pass through channel h ' ₁, h ' ₂... h ' _m-1send; Wherein, described channel h ' ₁, h ' ₂... h ' _m-1if, h ' ₁=h ' ₂=...=h ' _m-1=1, in the 2nd time slot, via node and each user's broadcast channel is exactly Rayleigh channel so;

If channel h ' ₁, h ' ₂... h ' _m-1be all to obey independent identically distributed rayleigh distributed, in the 2nd time slot, via node and each user's broadcast channel is exactly Rayleigh channel so;

Step 2, in all M-1 user, the broadcast message r receiving ₁, r ₂..., r _m-1adjudicate into wherein, r _irepresentative in second time slot, i the signal that user receives, r _i(1≤i≤M-1) is

$r_i=h_i^{\′}\·{\hat{s}}_R+w_i---\left(2\right)$

Wherein, w _i(1≤i≤M-1) is independent identically distributed white Gaussian noise, w _i, r _ibe n-dimensional vector;

be illustrated in second time slot i the signal r that user receives _iresult after being judged;

Step 3, calculates h ' _irepresent second time slot by via node to i subscriber channel;

Step 4, judgement whether all at R _userin, if so, carrying out step 5, second time slot finishes; Otherwise, be adjusted to step 1, will form by via node with broadcast resends;

Step 5, on M-1 user node respectively by unique decodable code encoder, shine upon, be decoded into time slot two finishes.

4. the relay transfer method based on unique decodable code according to claim 3, is characterized in that described unique decodable code coding is specially:

In the relay system of M node, suppose that the unique decodable code word set for i user (1≤i≤M-1) is C _i, and suppose, in each user's codeword set, to have | C _i| individual code word, each code word size is n, from the definition of unique decodable code, in i user, appoints and gets two different code word u _iand u ' _i, need to meet

${\mathrm{\Σ}}_{i=1}^{M-1}{u}_i\≠{\mathrm{\Σ}}_{i=1}^{M-1}{u_i}^{\′}---\left(3\right)$

Wherein, formula (1) two ends are all n-dimensional vectors, and vector is M system;

Order set E, for the step-by-step of all possible length of M-1 the user node code word that is n do and set, gather E and be expressed as i=1 wherein, 2 ..., M-1, according to the definition of unique decodable code, has in known set E individual element, the M system vector that each element is n dimension;

As follows in decode procedure, due to the character of M-1 user's unique decodable code, the known code word of M-1 arbitrarily and, just can be mapped to uniquely this M-1 code word, the decoded mode that is unique decodable code is according to code word and vectorial, by mapping, restores M-1 code word.

5. the relay transfer method based on unique decodable code according to claim 2, is characterized in that modulation system in the execution step two of described first time slot if BPSK, so

s _i＝2·d _i-I,(i＝1,2,...,M-1) (4)

Wherein, I be n tie up complete 1 vector be I=(1,1 ..., 1).

6. the relay transfer method based on unique decodable code according to claim 4, is characterized in that described judgement is hard decision or soft-decision.