Yung-Fa Huang|John Jenq|Cing-Chia Lai|Jyu-Wei Wang

Abstract—In future communications,cooperative communications with relay networks will be one of the most effective schemes to enlarge the coverage area and to boost the data rate.In the recent research results,the path selection,power allocation,and relay protocols on relay networks are the most important factors to improve the system performance.However,the channel quality of the direct transmission path and the relaying path has an influential effect on the performance of relay networks.Therefore,in this paper,we propose a best relaying path selection(BRPS)scheme to obtain the path diversity to improve the system capacity and data rate for cooperative networks (CNs).Simulation results show that the more the relay nodes are selected,the lower the bit error rate(BER)is.The proposed BRPS scheme obtains a high concession between both BER and system capacity for CNs.

1.Introduction

In the age of seamless wireless communications,the crammed full communications systems exhibit the interactivities in nowadays[1].The interactivities between the communications terminals become more important.Therefore,the design of the cooperative communications is the involution scheme of the next generation communications networks[2],[3].The multiple input and multiple output (MIMO) communications systems can effectively lower the bit error rate(BER)and improve the system capacity with multiple antennas[4]-[8].The multi-radio and multi-channel (MRMC) technique with cooperative communications is exploited to combat co-channel interference and improve the performance of multi-hop wireless networks[9].Cooperative communications or distributed MIMO systems are an emerging paradigm for increasing the spectral efficiency,coverage extension,and small cell deployment[10].The effect of polarization on the downlink achievable sum-rate of multi-user cooperative multi-point systems was studied using synchronous multi-link channel measurement with two different antenna arrangements[11].MIMO with a zero-forcing based linear receiver in the source-to-relay link was employed to enhance the radio frequency(RF)link data rate[12].The amplify-and-forward(AAF)based cooperation in a MIMO relay network with two antennas at each node over the independent but not necessarily identically distributed Nakagami-m fading channels system has been studied in[13].However,the multiple antennas for effective MIMO are not suitable to equip in the small handset.In cooperative networks (CNs),the path diversity increases the coverage of wireless communications systems[14].Moreover,the spatial diversity increases the capacity of the wireless channels[15].Thus,the relay nodes can relay data from the source node to the destination node to increase the path diversity.

2.System Model

In the cooperative communications,data are modulated by binary phase shift keying(BPSK)as shown in Fig.1[16].Moreover,the networking model for CNs consists of three parts:One source nodeS,one destination nodeD,and several relay nodesRi,as shown in Fig.2.As for the channels,there are three kinds of channels:Source to relay,hs,ri,source to destination,hs,d,and relay to destination,hri,d.

Fig.1.System model for the wireless communications systems.

Fig.2.Network model for the cooperative communications.

In the wireless communications networks,the transmitted electromagnetic signal suffers the path loss and fading induced by the multiple reflections,refractions,and scattering through the obstacles.Thus the receivednth symbol signal at the receiver of destination is obtained by

wherexs={+1,–1} is the transmitted data,as,dis Rayleigh fading,andds,dis the path loss from sourceSto destinationD.Thus,hs,d=ds,das,d.The termzs,dis the added white Gaussian noise(AWGN)with zero mean and varianceσ2for the link of the source node to the destination node.

2.1.Cooperative Models

In the basic cooperative communications,there are three nodes and three wireless channels.When the source node transmits data to the destination node,there are three types to perform the transmission.The first one is that the source node transmits data directly to the destination node through the direct path,hs,d.Then the received signal is obtained by

The second one is that the source node transmits data to the relay node.Thus,the received signal at therith relay node can be obtained by

where the termzs,ris AWGN with zero mean and varianceσ2for all links of the source node to relay nodes.And then the relay nodes relay data to the destination node.Thus,the received signal at the destination node is obtained by

The third is called as the cooperative mode which combines the two signals received from the source directly and from the relay nodes by relaying[16].The termzr,dis AWGN with zero mean and varianceσ2for all links of the relay nodes to the destination node.

2.2.Cooperative Protocols

Moreover,to obtain the effective cooperation,two types of cooperative protocols have been developed[16],[17].AAF[16]can be used to simply amplify the relaying signal while the computation power of the relay nodes is limited.However,there exists a weakness of the noise enhancement.Thus,to compromise the signal amplification and noise enhancement,the amplified gain can be optimized by

wherehs,ris the estimated channels gain[16],ξ=E[|x|2]is signal power (E[·]represents the expectation operation),andis the variance of noise in the channel from the source node to the relay nodes.The other scheme is the decode-and-forward(DAF)[3],which relays the signal after decoding the received signal correctly.

2.3.Combining Techniques for CNs

In the wireless CNs,the received signal at the destination node would probably be obtained from more than two paths.Then many combining methods can be selected to optimize the received signal to noise ratio(SNR).

The selection combining (SC)is the simplest method to select the most adequate channel to transmit the signal.The receiver needs to continually measure the channels situation to provide a good choice.The received signal by SC is obtained by

where the symbol ∠hs,din the superscript represents the phase part of the linkhs,d.SNRs,r,drepresents SNR of the link of source-relay-destination.To maximize SNR,the maximum ratio combining(MRC)[16]is performed by weighting the signal power of each path.When MRC is applied to the three nodes model,the received signal is obtained by

whereg1andg2are the weights applied to the direct path and the relaying path,respectively,by

for the DAF scheme.If the relaying path is with the AAF scheme,g1andg2are obtained by

where[·]*is the conjugate operation.

2.4.Link Quality Estimations

To perform the adaptive cooperative communications,the channel side information should be estimated and link quality should be evaluated[16].When BPSK is used,BER for the single link quality based on channel side information(CSI)can be obtained by

wherehis the channel gain,γis the instantaneous SNR of the channel,andQ(x)is the complementary error function given by[18]:

Similarly,when QPSK is used,the instantaneous BER for the single link quality based on CSI can be obtained:

However,in the cooperative communications,the link quality should be estimated based on the cooperative communications link.That is,with the exception of the direct link,the links of the relaying paths should also be included for the link estimation.Therefore,after BERs for CNs(S,R,andD)are gathered,with the BPSK scheme,the instantaneous SNR of the cooperative link can be estimated by

where BERs,r,dis BER of the cooperative link in CNs(S,R,andD).Similarly,with QPSK,the instantaneous SNR of the channel can be estimated by

2.5.Channel Capacity for CNs

The channel capacity depends on the channels impulse response and SNR.Thus the channel capacity in a direct transmission path can be obtained by

whereEb/Nois SNR in the channel.Moreover,the channel capacity in a DAF scheme can be obtained by[17]

where the fraction 1/2 means that two time slots are used in the path.Therefore,it is easy to know that the relaying path decreases the system capacity in the cooperative communications.Besides,to perform effective relaying for the DAF scheme,the error coding should be included in the transmission of the source node to the relay nodes to ensure the decoding decision.In this paper,we add cyclic redundancy codes(CRCs)to perform the data error checking before the data relaying at the relay node.

3.Channel Capacity Based Relaying Path Selection

In the cooperative communications,CSI is assumed to be perfectly estimated,such that the combining schemes can be used to perform the diversity gain.Moreover,the cooperative schemes of SC and MRC can be selected to obtain the required capacity.The optimization for CNs is obtained with SC and MRC via DAF protocols[3]in which the best relay selection method is shown as Fig.3.

Fig.3.System model of relaying networks for the proposed best relay selection scheme with the number of relay nodes I=2.

However,in this paper,the optimal cooperative path is obtained with SC by

Table 1:Instantaneous SNR for the proposed APS schemes

where

Moreover,with the perfect channel estimation,an adaptive threshold of the instantaneous channel gain is proposed to select the highest quality path.

4.Simulation Results

In order to verify the performance of CNs,we performed the computer simulation using MATLAB programming for the system models in Fig.3.The system models and simulation parameters are shown in Table 2.

At first we investigated the performance of the MRC receiver for CNs as shown in Fig.4.It is observed that MRC can improve the BER performance.The non-selection scheme provides more diversities of the cooperative links for CNs.

Moreover,the channel capacity in CNs is compared with the number of relay nodesI=1 andI=2,as shown in Fig.5.It is observed that when the number of relay nodesIincreases,the system capacity declines.That is because with TDMA,the channel capacity will be decreased by 1/(I+1).Moreover,the performance with CRC in the dash line slightly increases the system capacity for the lower SNR environment.

Table 2:Simulation structure and system parameters

Fig.4.Comparisons of the selection on the best path for I =1 and I =2 with the MRC scheme for CNs.

Moreover,we investigated the performance of the MRC receiver for CNs as shown in Fig.6.From Fig.6,it is observed that MRC improves the BER performance.The multiple-input single-output (MISO)systems with two transmitter antennas and one receiver antenna denoted by 2×1 can obtain the maximal two-path diversity gain.The non-selection scheme withI=2 MRC provides more diversities of the cooperative links for CNs.However,the proposed BRPS scheme approaches the best performance comparing with the other cooperative communications schemes.

Fig.5.Capacity of CNs for I =1 and I =2 with CRC.

Fig.7.Capacity of CNs for I=1 and I=2.

The channel capacity in CNs was compared with the number of relay nodesI=1 andI=2,as shown in Fig.7.From Fig.7,it is observed that as the number of relay nodesIincreases,the system capacity declines.That is because with TDMA,the channel capacity will be decreased by 1/(I+1).

Furthermore,to investigate the need of the number of relay nodes,the average number of relay nodes is calculated for the comparisons as shown in Figs.8 and 9.From Figs.8 and 9,the proposed BRPS scheme needs less relay nodes than the conventional DAF schemes.

Fig.8.Comparison of the number of relay nodes for I=1 with MRC.

Fig.9.Comparison of the number of relay nodes for I=2 with MRC.

Therefore,it can be derived that both BER and capacity performance are exclusive with each other.The path diversity gain with the relay nodes improves the BER performance but lowers the system capacity as shown in Fig.10.

Fig.10.Compromise on both BER and system capacity for CNs.

5.Conclusions

In this paper,we investigated the relay diversity gain for CNs.Moreover,we proposed a BRPS scheme to improve the system capacity and data rate for CNs.Simulation results showed that the more the relay nodes were selected,the lower BER.The proposed BRPS could obtain the compromise between BER and system capacity for CNs.