Search

Nowadays, Leakage detection is of great importance as pipelines are the major means of transporting hydrocarbon fluids and gases. In this paper, a novel two-stage detection method is introduced to differentiate normal, leakage and transient conditions of pipelines. In this method, feature vectors are constructed from the flow difference and pressure using leakage characteristics, and are normalized with the modified hyperbolic-tangent estimator. An artificial neural network is used in the first stage of detection to differentiate normal and abnormal conditions with the feature vectors as the inputs. In the second detection stage, a simple logic is used to distinguish leakage and transient from abnormal time-windows. In addition, a pre-set leak-size tolerance is used to trigger alarms for detected leakage time-windows. The results for the cases of using different machine learning methods and varying leak-size tolerances are given. The method has been shown to have higher detection performance and less false alarms in comparison with the line balance and Kantorovich distance methods.

In this paper, a statistical-based sphere decoding with increasing radius search (S-SD-IRS) algorithm is proposed, where the radiuses of the decoding hyperspheres are determined based on the statistical properties of the communication channel and additive noise. We show that the probability density functions (PDFs) of the q lowest squared distances in the closest lattice point problem can be approximated by Gumbel distributions with different parameters. Based on the obtained PDFs and by considering the characteristics of the fading channels and additive noise, we choose the radiuses for sphere decoding more efficiently than the conventional methods that ignore the characteristics of system. The performance achieved by the proposed algorithm is very close to the optimal maximum likelihood decoding (MLD) over a wide range of signal-to-noise ratios (SNRs), while the computational complexity, compared to existing sphere decoding variants, is significantly reduced. It is shown that the average number of lattice points inside the decoding hyperspheres drastically reduces in the proposed S-SD-IRS algorithm.

This paper investigates the power consumption and user-clustering of a multi-cluster multi-input single-output non-orthogonal multiple access (MISO-NOMA) systems. First, we derive the average transmit power of a two-user cluster, given constraints on the signal-to-interference-plus-noise-ratios (SINRs) of users and a threshold on the alignment of channel directions. Then the asymptotic behavior of the average transmit power under several scenarios are studied to show the scaling of the transmit power with respect to the base station (BS) antenna number and the alignment threshold. After that, a two-layer user clustering algorithm is proposed to improve the power efficiency by selecting the head and tail users in each cluster sequentially. Simulation results validate the correctness of our theoretical results. Further, the proposed user-cluster algorithm is shown to achieve significant saving on the transmit power.

In this paper, a decision-directed (DD)-channel estimation (CE) algorithm by employing deep learning (DL) is proposed for high mobility vehicular environments. The proposed algorithm relies on DL for channel prediction without prior knowledge about channel statistics, such as the channel covariance matrix and its time variation. Therefore, it does not require any prior Doppler rate estimation, which is highly complicated in high mobility environments. Based on the performed simulations, comparing to the Kalman filter-based DD-CE algorithm, our DL-based algorithm results in more reliable communications. It has been shown that comparing to the minimum mean square error (MMSE)-based DD-CE algorithm, our DL-based algorithm imposes much lower complexity to the system and the performance degradation is small compared to the MMSE-based DD-CE algorithm that requires the exact value of the Doppler rate.

To support reliable transmission of data at high rate in time-varying fading channels, adaptive transmission is required, where transmitter and receiver adjust their transmission and reception mode to the dynamics of the channel. The receiver, based on its channel estimation and prediction, decides the optimal link adaptation and feeds this back to the transmitter. In this paper, we develop a deep learning (DL)-based link adaptation algorithm for highly dynamic communication links, where adaptive transmission parameters are decided for $l\gt1$ forward time steps without a priori knowledge on channel statistics. Compared to conventional solutions, our approach reduces the feedback requirements from the receiver to the transmitter by a factor of l which significantly reduces the complexity. This achievement comes at no additional cost on the data rate and/or bit error rate.

This paper investigates the abnormality detection based on the Renyi divergence with the divergence order between 0 and 1. It is assumed that univariate Gaussian data samples are independent and identically distributed (i.i.d.). The distance between the estimated distribution from observed data and that from historical data under the normal condition is quantified by the Renyi divergence, the proposed test statistic. The false alarm rate (FAR) and the missed alarm rate (MAR) are derived analytically based on the distribution of the Renyi divergence. Under the abnormal condition, constant bias and multiplicative faults are considered. Taking both the FAR and MAR into consideration, the proposed detection algorithm can adaptively optimize the divergence order and threshold according to observed data. In the simulation, the analytical FAR and MAR results are verified by the Monte Carlo simulations, and the proposed algorithm is shown to outperform the Kullback-Leibler divergence based method.

As a key infrastructure of Internet of Things (loT), wireless sensor networks (WSN) can be utilized in a wide range of applications. The prediction based data fusion methods provide effective tools to reduce the amount of data transmissions while maintaining prediction accuracy. Recently a grey prediction model (GM) combining optimally-pruned extreme learning machine (OPELM) data fusion method has been proposed and shown to have good performance. However, the existing GM- OPELM method performs model training and broadcasting before each prediction, resulting in high complexity and energy consumption. In this paper the conditional training based GM (CT-GM) and GM-OPELM (CT-GM-OPELM) are proposed. By introducing an error threshold, the algorithms only perform model training when the prediction error is beyond the threshold. Compared with existing GM and GM-OPELM methods, the CT- GM and CT-GM-OPELM methods not only can achieve the higher rate of acceptable prediction and better time efficiency but also has significant reduction in the energy consumption on model training and transmissions.

This paper is on the uplink communications of single-cell multi-user massive multi-input-multi-output (MIMO) systems with a mixture of 1-bit analog-to-digital converters (ADCs) and full-resolution ADCs. A new low-complexity linear combining scheme is proposed for the receiver at the base station (BS) to take into account the difference of the two ADC types. A symbol-wise decoding is derived to separately decode the information of each user and an analytical approximation is obtained for the symbol error rate (SER) of the proposed combining and decoding. Simulation results show that the proposed scheme achieves significant performance improvement compared to the traditional maximum-ratio-combining (MRC) receiver.

This paper considers a multiple-input multiple-output (MIMO) multi-way relay network (MWRN) where users exchange their information via a multi-way relay equipped with a large-scale antenna array, i.e., massive MIMO multi-way relay. Further, each antenna at the relay station is assumed to have a pair of low-resolution analog-to-digital converters (ADCs) to reduce the energy consumption and hardware cost at the relay. Lloyd-max algorithm is used to find the mean-squared error (MSE) optimum quantization labels and thresholds for the ADCs. With perfect channel state information (CSI) and zero-forcing (ZF) beam-forming for both reception and transmission at the relay, a closed-form approximation for the average achievable rate of each pair of users is derived with the help of Bussgang's decomposition. The results enable us to understand the achievable rate behavior with respect to system parameters, and especially to quantify the performance degradation caused by low-resolution ADCs. Numerical results verify the validity of Bussgang's theorem in our case, and that the derived result is an accurate performance predictor of the network. Further, both analytical and theoretical results reveal that the effect of ADC resolutions on the rate performance is as significant as the relay and users' transmit powers.

This paper investigated the uplink of multi-user massive multi-input multi-output (MIMO) systems with a mixed analog-to-digital converter (ADC) receiver architecture, in which some antennas are equipped with costly full-resolution ADCs and others with less expensive low-resolution ADCs. A closed-form approximation of the achievable spectral efficiency (SE) with the maximum-ratio combining (MRC) detector is derived. Based on this approximated result, the effects of the number of base station antennas, the transmit power, the proportion of full-resolution ADCs in the mixed-ADC structure, and the number of quantization bits of the low-resolution ADCs are revealed. Results showcase that the achievable SE increases with the number of BS antennas and quantization bits, and it converges to a saturated value in the high user power regime or the full ADC resolution case. Most important, this work efficiency verifies that for massive MIMO, the mixed-ADC receiver with a small fraction of full-resolution ADCs can have comparable SE performance with the receiver with all full-resolution ADCs but at a considerably lower hardware cost., Comment: 11 pages, 3 figures

This paper studies the relay selection (RS) problem for full-duplex (FD) relay networks with multiple source-destination (SD) pairs under the attack of colluding eavesdroppers. Based on available channel state information (CSI), both optimal relay selection (ORS) and suboptimal relay selection (SRS) schemes are considered to maximize the minimum secrecy rate among all pairs in order to enhance the physical-layer security. The secrecy performance of the more practical SRS scheme is then evaluated in terms of intercept probability and diversity order. The SRS achieves full diversity when the gains of the main-to- eavesdropper and the main-to-interference channels increase asymptotically.

Zero-forcing (ZF) precoding plays an important role for massive MIMO downlink due to its near optimal performance in high signal-to- noise (SNR) region. However, the high computation cost of the involved matrix inversion hinders its application in practical large- scale systems. In this paper, we adopt the first order Neumann series (NS) expansion for a low-complexity approximation of matrix inversion. Compared to existing NS based schemes, we introduce a relaxation parameter jointly with one user's channel interference to others into the precondition matrix and propose the identity-plus- column NS (ICNS) method. By further exploiting the multi-user diversity gain via choosing the user with largest interference to others, the ordered ICNS method is also proposed. Moreover, the closed-form sum-rate approximation of the ICNS method is derived. Simulations verify our analytical results and the advantage of the proposed schemes over other existing low-complexity ZF precodings for massive MIMO systems with correlated channels and not-so-small loading factor.

In this paper, we study the beam-based training design for the single-user (SU) hybrid massive antenna system based on outage probability performance. First, an interleaved training design is proposed where the feedback is concatenated with the training procedure to monitor the training status and to have the training length adaptive to the channel realization. Then, the average training length and outage probability are derived for the proposed interleaved training and SU transmission. Analytical results and simulations show that the proposed interleaved scheme achieves the same outage performance as the traditional full-training scheme but with significant saving in the training overhead.

This work provides a comprehensive scaling law based performance analysis for single-cell multi-user massive multiple-input-multiple-output (MIMO) downlink systems with maximal-ratio transmission (MRT). Channel state information (CSI) error and channel spatial correlation are both considered. First, a sum-rate lower bound is derived by using Jensen's inequality and the asymptotically deterministic property of the received signal power. Via a general scaling model on the number of users, the CSI quality and the data transmission power, with respect to the number of base station (BS) antennas, the asymptotic performance scaling law is obtained, which reveals quantitatively the tradeoff of the above three parameters. In addition, the applicability of the derived asymptotic scaling law in practical systems with not-so-large antenna numbers is discussed. Finally, all analytical results are evaluated through simulations.

This paper conducts the performance sensitivity analysis for linear alarm filters. With a derivative-based local sensitivity measure, models are proposed to assess the sensitivity of the system detection errors to changes in the trip point and uncertainties in the collected data. Then, sensitivity formulations for a general linear alarm filter framework with unknown data distributions are studied. Extensive simulations with Gaussian and Gamma data using two typical linear filters, the moving average (MA) filter and the linear weighted (LW) filter, are conducted to verify the analytical results and to show the trade-offs among different system parameters.

This work analyses the performance of multi-user massive MIMO relay networks, where the relay station is equipped with a large number of antennas. The combination of maximal-ratio-combining (MRC) and maximal-ratio-transmission (MRT) is utilized at the relay for low-complexity processing. Different from existing work, we consider the practical scenario with limited number of users, fixed source power, and fixed relay power. Our work shows that the interference is neither negligible nor asymptotically deterministic, and it dominates the statistical properties of the signal-to-interference-plus-noise ratio (SINR). Via deriving an approximation on the probability density function (PDF) of the interference power, analytical results on the sum-rate and outage probability performance of the multi-user relay network are obtained.

For single-cell multi-user massive multi-input-multi-output (MIMO) system downlink, we consider the per-antenna power constraint and propose a modified maximum-ratio-transmission (MRT) precoding scheme. The outage probability performance of the scheme are analyzed. To enable the analysis, we study the random behaviour of the signal-to-interference-plus-noise-ratio (SINR) and derive an approximation for the probability density function (pdf) of the interference power. Simulation results are shown to validate the theoretical derivations. Moreover, our work show that the massive MIMO system with per-antenna power constraint under the modified MRT precoding can achieve lower outage probability than that with a total power constraint under the standard MRT precoding, even though its power constraint is more strict.

This paper proposes a low-complexity precoding scheme for the downlink of single-cell two-user massive multi-input-multi-output (MIMO) systems, which can be seen as a combination of matched-filter (MF) precoding with extra user power adjustment. In this scheme, the base station (BS) antennas are grouped into two sets according to their channels, then each set has a main user to serve and a power coefficient is introduced to adjust the weighting between main user and the other user. Via theoretical analysis on the statistical properties of the signal power and interference power, the achievable sum-rate is derived. In addition, the power adjustment coefficient optimization with respect to the sum-rate maximization is solved. Our results show that the proposed precoding outperforms the standard MF precoding.

In practical multi-input-multi-output (MIMO) systems, the channel matrices often have reduced rank especially for propagation environments with limited scattering. For this scenario, reliable detection of the channel rank is essential in achieving the high performance provided by the MIMO configuration. In traditional rank detection schemes, a single threshold is used and the channel rank is detected as the number of singular values of the observation matrix that are larger than the threshold. In this work, we propose a multiple-threshold-based rank detection scheme, where each threshold corresponds to one possible rank value. The thresholds are optimized based on the derived lower bounds on the rank detection probability for different channel rank values. Simulation results show that the proposed detection scheme outperforms existing schemes in the probability of correct rank detection.

For a multi-relay amplify-and-forward (AF) network, an improved training scheme is proposed and the corresponding training power allocation problem is investigated for the individual channel estimation at the destination. Compared with a previous scheme where the training of the source-relay channels is conducted sequentially, the proposed scheme exploits the broadcast nature of the source-relay links to save training time and power. By analyzing the total mean square error (MSE) of all channel estimates and with the help of approximations, we derive a training power allocation solution among the source and the relays. Simulation results demonstrate that our improved training scheme and power allocation solution obtain lower total MSE than the existing schemes.

The ever increasing demands for mobile Internet access have resulted a significant growth in bandwidth usage. Multi-way relay networks (MWRNs) can address this challenge by improving the system spectral efficiency. In this work, we further improve the spectral efficiency of MWRNs by proposing a novel linear beamforming design, namely partial zero-forcing (PZF), for MWRNs with a multiple-input-multiple-output relay. Compared to zero-forcing (ZF), in PZF, the constraints on the relay beamforming are relaxed such that only partial user-interference, instead of all, is canceled at the relay. After receiving the relay's signal, the users eliminate the remaining interferences through self-interference and successive interference cancellation. By taking such two-stage interference cancellation, PZF better exploits the degrees-of-freedom in MWRNs and achieves higher data rates than ZF.

In this paper we study a cellular network, in which multiple-antenna users communicate with multiple-antenna base stations (BSs) through fixed infrastructure-based multiple antenna relay stations (RSs). With cooperation among the BSs and linear processing at the RSs, we aim to find the optimal precoding matrices at the users and the beamforming matrices at the RSs that jointly maximize the system sum rate. Unlike for the conventional uplink (without relays), the sum-rate optimization is non-convex. More so, there is cross-coupling of the RSs' channels due to the forwarded interferences by the RSs. Firstly, we incorporate interference pre-cancelation into the RSs' beamforming designs. Secondly, we match each RS's beamforming matrix to the corresponding backward and forward channels such that the end-to-end channel is diagonalizable. Furthermore, we propose an iterative alternating minimization based algorithm to maximize the system sum rate. Finally, we consider two user-RS scheduling/mapping schemes namely the “random” and “channel-aware” schemes. Simulation results show that the channel-aware scheme outperform the random scheme with the performance gap unchanged with increasing number of antennas at the nodes.

In this paper, we adopt a novel efficiency measure, the power-normalized SNR (PN-SNR), in the design of multi-relay networks with an individual power constraint for each relay. We formulate the PN-SNR maximization problem with a received SNR constraint to maintain the quality of service (QoS) and propose a low complexity algorithm for a suboptimal solution. The average PN-SNR and average throughput of the proposed design are simulated and compared with existing designs. We also simulate the probability of no transmission with the proposed design for an arbitrary SNR level. We observe that the proposed design is more power efficient than existing ones with comparable performance in average throughput. When the transmitter power is moderate, with higher QoS constraint, the probability of no transmission gets higher; the average throughput decreases, but the PN-SNR of the network increases.

For a three-node amplify-and-forward (AF) relay network, we investigate the power allocation problem during the channel training process for the destination to estimate the global channel state information (CSI) of the whole network. Linear minimum-mean-square-error (LMMSE) estimation is adopted. The mean-square-error (MSE) of the channel estimation is analyzed, which shows that the quality of the source-to-relay channel estimation depends on the quality of the relay-to-destination channel and its estimation. The outage probability (OP) of the network with channel estimation error is also calculated. Then power allocations between the source and the relay for different training steps are derived based on the total MSE and the OP. The performance of the proposed power allocations is simulated and shown to be superior to even power allocation.

For a multi-user multi-relay network with a total relay power constraint, we investigate the relay power allocation (PA) that maximizes the network sum-rate. Different users use orthogonal channels to avoid interference. With the optimal relay beamforming for each user, the problem reduces to finding the optimal power the relays use in total to help each user. We first prove that the problem is convex. Then a suboptimal solution is proposed in closed form. Further, the SNR and sum-rate are analyzed for networks with a large number of relay antennas. The asymptotic behaviour of the SNR is derived rigorously for the high transmit power regime. Simulation results are provided to show the significance of proper PA and to justify the analytical SNR and sum-rate results.

For a three-node amplify-and-forward (AF) two-way relay network (TWRN), we investigate the power allocation problem during the channel training process for both terminals to estimate the global channel state information (CSI) of the whole network. This includes the power allocations between different training steps and among different nodes. Two power allocation schemes are proposed: a mean-square-error (MSE) based scheme and an outage probability (OP) based scheme. Our simulation results demonstrate that the proposed power allocations are superior to even power allocation.

In this paper, we adopt a novel efficiency measure, namely, the received signal to noise ratio (SNR) per unit power, in relay network design. First, limitations of conventional efficiency measures, spectral efficiency and energy efficiency, are discussed to motivate the SNR-per-unit-power (SNR-PUP) measure. Then for a single-relay network which uses amplify-and-forward (AF) protocol, we find the optimal relay power that maximizes the SNR-PUP for a given transmitter power. The average relay power, the SNR-PUP, and the outage probability of the proposed design are investigated analytically and numerically, and are compared with the conventional design where the relay power is fixed. We also consider a general multi-relay network and use gradient-ascent method for the SNR-PUP maximization. Our results show that with the same average relay transmit power, the proposed design is superior not only in the SNR-PUP but also in the outage probability for both single and multi-relay networks.

For bidirectional multi-relay networks, where each node is assumed to have its own power constraint, the optimal relay and user power control is considered via maximizing the smaller signal-to-noise ratio (SNR) of the two end users in the network. A numerical algorithm is proposed to find the optimal solution. Then, a suboptimal algorithm with low complexity is proposed. The performance of this algorithm is shown numerically to be close to the optimal. It is also shown via simulation that proper user and relay power control significantly improves the network performance, especially when the power constraints of the two end-users in the networks are unbalanced.

For a general multi-input-multi-output (MIMO) relay network, an estimation method for the receiver to obtain the end-to-end channels is proposed. Instead of straightforwardly estimating entries of the end-to-end channel matrix, the proposed scheme takes into consideration the special structure of the end-to-end channel matrix. By parameterizing the channel matrix with its singular values and singular vectors using singular value decomposition (SVD), the proposed scheme estimates the singular values and left and right singular vectors, which are then combined to form an estimation of the overall channel matrix. The proposed estimation follows the maximum-likelihood (ML) estimation method. Simulations on the mean square error (MSE) of the channel estimation are presented, which show the advantage of the proposed scheme over straightforward estimation of the channel entries for networks whose transmitter and receiver are equipped with multiple antennas.

In this paper, we consider a multi-source single-relay wireless network, where the relay facilitates transmissions of the sources' signals to the destination. We study the relay power allocation among the sources, and use the bargaining theory to model the negotiation among the sources on fair allocation of the relay power. By assigning a bargaining power to each source to indicate its transmission priority, we propose an asymmetric Nash bargaining solution (NBS)-based relay power allocation scheme. The impact of the bargaining powers on the relay power allocation and network performance is analyzed. We show that the proposed scheme addresses the tradeoff between the sum-rate of the network and the fairness among the sources, and can be adapted to meet different requirements in different applications by proper selection of the bargaining powers.

This paper considers a multi-user single-relay wireless network, where the relay gets paid for helping the users forward signals, and the users pay to receive the relay service. We study the relay power allocation and pricing problem, and model the interaction between the users and the relay as a two-level Stackelberg game. In this game, the relay, modeled as the service provider and the leader of the game, sets the relay price to maximize its revenue; while the users are modeled as customers and the follower who buy power from the relay. For the relay power allocation among users, we use a bargaining game model to achieve a fair allocation. Based on the proposed fair relay power allocation rule, we then analyze the optimal relay power price that maximizes the relay's revenue, and derive the analytical solution. Simulation shows that the proposed power allocation scheme achieves a higher network sum-rate than the even power allocation, and is fairer than the sum-rate-optimal allocation. We also show that the proposed pricing and power allocation solution is consistent with the laws of supply and demand.

In this paper, channel training and coherent decodings with channel estimation errors are investigated in a relay network with one single-antenna transmitter, two single-antenna relays, and one double-antenna receiver. A two-stage training scheme is proposed to estimate both the relay-receiver and transmitter-relay channels at the receiver. We use distributed space-time coding (DSTC) for data transmission and investigate the effect of channel estimation errors on the diversity. Two coherent decodings are considered: mismatched decoding in which the channel estimations are treated as if perfect, and matched decoding which takes into account the estimation errors. We show that for full diversity, mismatched decoding requires at least 6 symbol intervals for training, while 4 symbol intervals for training are enough for matched decoding. On the other hand, the complexity of matched decoding is much higher. To achieve a balance between performance and complexity, an adaptive decoding scheme is proposed. Simulated network error rates are shown to justify the analytical results.

This paper studies a multiple-input multiple-output (MIMO) broadcast relay channel (BRC) in which a multiple-antenna base station (BS) communicates with multiple-antenna users through a fixed infrastructure-based multiple-antenna relay station (RS). Applying dirty paper coding (DPC) at the BS and linear processing at the RS, our goal is to find the optimal input covariance matrices at the BS and the beamforming matrix at the RS that jointly maximize the system sum-rate. To solve the optimization problem, a more tractable dual multiple access relay channel (MARC) is investigated and an iterative algorithm is proposed to obtain the optimal matrices for the dual system. The mapping from the resulting covariance matrices for the MARC to the covariance matrices for the original BRC is derived. Unlike other existing MIMO BRC schemes, designed for networks with single-antenna users only, our solution is applicable to networks with multiple-antenna users. Compared with two such single-antenna-user schemes, simulations show that the proposed scheme outperforms the all-pass relay design and performs similar to the SVD-relay design. Also, the proposed design performs close to a sum-rate upper bound with the gap decreasing with increasing number of users' antennas.

This paper considers multi-access relay networks (MARNs) where two single-antenna users communicate to one N-antenna receiver via two hops of transmissions through one R-antenna relay. Nodes in the network are under two linear constraints. The relay linearly maps its received signal to generate its forwarding signal without decoding; the receiver has linear decoding complexity in the number of users. Since the relay-receiver link has more transmission paths than the link from each user to the relay, a two-step protocol, called TDMA-ICRec, is proposed. In the first step, both users timeshare the user-relay link. Linear combining is then performed at the relay to maximize the signal-to-noise-ratio (SNR) for each user. In the second step, the relay forwards both users' symbols concurrently to the receiver to enhance the transmission rate. At the receiver, an interference cancellation (IC) technique is used to decouple the users and ML decoding is conducted to recover each user's symbols. Two network scenarios are studied when the relay has two antennas and four antennas. Through analysis and simulation, when the receiver has more than two antennas, TDMA-ICRec achieves the same diversity as the full-TDMA-DSTC interference-free (int-free) scheme, yet with higher symbol rate.

This paper investigates the channel training and estimation problems for distributed space-time coding (DSTC). To use DSTC in multi-antenna relay networks, the receiver needs to know the channels between the relays and the receiver (Relay-R channels) and also the equivalent channels between the transmitter and the receiver (Equ-T-R channels). By sending pilot signals from the relays, the training of the Relay-R channels is equivalent to that of a multi-input-multi-output (MIMO) system. The training of the Equ-T-R channels can be conducted directly using DSTC; but it requires a long training period. Thus, a separate-training method is proposed, in which an estimation on the Equ-T-R channels is obtained from estimations on the channels from the transmitter to the relays (T-Relay channels) and the Relay-R channels. The pilot code designs that minimize the trace of the error covariance matrix are investigated. The requirements on the training period are also derived, from which an adaptive training-period design is proposed. Simulation shows that in some networks, even with shorter training period, the separate-training scheme can achieve better performance than the direct-training scheme.

This paper is on interference cancellation (IC) schemes for a two-user relay network where users are allowed to communicate simultaneously. The considered networks have one double-antenna half-duplex relay, single-antenna receiver but three scenarios on both users: single-antenna, double-antenna, and four-antenna. We apply the IC scheme, which was originally proposed for multi-antenna multi-user direct communication systems, to multi-user relay networks and propose a protocol called IC-Relay-TDMA, in which the relay cancels user interference and then amplifies and forwards the interference-free (int-free) user information to the receiver in TDMA. The maximum likelihood (ML) decoding at the receiver can be conducted symbol by symbol for both networks with single-antenna users and double-antenna users. Compared to the full TDMA scheme in which each user is allocated different time slots from end to end to avoid interference, IC-Relay-TDMA achieves the same diversity with a higher symbol rate when both users have two or four antennas.

Distributed space-time coding achieves full diversity in wireless relay networks without channel information at the relays. This paper is on the design of specific distributed space-time codes. By Cayley transform, the design of unitary matrices is transformed to that of Hermitian matrices. Gradient-descent method is used in the optimization to find codes that are “close” to orthogonal designs, which is known to be optimal in both coding gain and diversity. The proposed code design method can be used for networks with any relays and transmission rate. The optimized codes are adaptive to the information symbol constellation. Simulation shows that the designed codes achieve full diversity, and outperform the random code and other existing distributed space-time codes.

Distributed space-time coding (DSTC) is a cooperative scheme proposed for wireless relay networks that achieves full diversity when the transmit power is infinitely high. In this paper, we propose to employ maximum-ratio combining (MRC) at the multiple-antenna relay to improve the performance of DSTC. Simulations and analysis show that the new scheme achieves a much better performance in relay networks than DSTC. In some network scenarios, it can even improve the scaling of the error rate with the transmit power. Furthermore, the new scheme requires a shorter training interval than DSTC.

A relay selection (RS) scheme for two-way networks with multiple amplify-and-forward relays is proposed. The scheme maximizes the worse receive signal-to-noise ratio (SNR) of the two end users. It is proved to coincide with the best-worse RS scheme originally proposed for one-way relay networks. Through rigorous error rate analysis, the scheme is shown to achieve full diversity. For networks with two relays and any power condition and constellation, an approximation on the block error rate (BLER) of the scheme is derived in closed-form. It is shown by simulation to be effective in estimating the network reliability in the high transmit power region.

This paper considers the interference cancellation (IC) problem in multi-user wireless relay networks. First, it is shown that using distributed space-time coding (DSTC), the multiple antenna IC scheme previously proposed for systems with direct transmissions can be applied to relay networks. The ML decoding after full IC can be performed symbol by symbol. Then, by allowing IC at relays, a new degree of freedom in relay network design is discovered. With this new idea, the required number of antennas at the receiver for full IC can be reduced and a balance between diversity and delay can be obtained.

This paper is on relay selection (RS) schemes for wireless relay networks. First, we derive the diversity of many single RS schemes in the literature. Then, we generalize the idea of RS by allowing more than one relay to cooperate. Several multiple RS schemes are proposed, which are proved to achieve full diversity. Simulation results show that they perform much better than the corresponding single RS methods and very close to the optimal multiple RS scheme. However, the computational complexity of the suboptimal schemes are linear in the number of relays, far superior to the optimal selection, which has exponential complexity. In addition, when the number of relays is large, the multiple RS schemes require the same amount of feedback bits from the receiver as single RS schemes.

This paper is on relay beamforming in wireless networks, in which the receiver has perfect information of all channels and each relay knows its own channels. Instead of the commonly used total power constraint on relays and the transmitter, we use a more practical assumption that every node in the network has its own power constraint. A two-step amplify-and-forward protocol with beamforming is used, in which the transmitter and relays are allowed to adaptively adjust their transmit power and directions according to available channel information. The optimal beamforming problem is solved analytically. The complexity of finding the exact solution is linear in the number of relays. Our results show that the transmitter should always use its maximal power and the optimal power used at a relay is not a binary function. It can take any value between zero and its maximum transmit power. Also, interestingly, this value depends on the quality of all other channels in addition to the relaypsilas own ones. Despite this coupling fact, distributive strategies are proposed in which, with the aid of a low-rate broadcast from the receiver, a relay needs only its own channel information to implement the optimal power control. Simulated performance shows that network beamforming achieves full diversity and outperforms other existing schemes.

This paper is on beamforming in networks whose relays know the channel means and covariances. The question we answer is: To optimize the network performance, how much power should each relay use? Instead of the widely used aggregate relay power constraint, we use the more practical assumption that each relay has a separate power constraint. We generalize the distributed space-time coding scheme so that each relay can adapt its transmit power according to the channel information. For two-relay networks, we analytically solve the relay power control problem at high transmit powers using the pairwise error probability (PEP) minimization. Simulation shows that appropriate relay power control can largely improve the reliability, especially when the qualities of relay paths are far apart.

Distributed space-time coding was proposed to achieve cooperative diversity in wireless relay networks, in which antennas of relays work as transmit antennas of the sender and generate a space-time code at the receiver in a distributive way. It was proved that this scheme achieves the maximum diversity R when the total transmit power is infinitely large, where R is the number of relays in the network. This paper is on the design of practical distributed space-time codes for wireless relay networks. We use orthogonal and quasi-orthogonal designs which are originally used in the design of space-time codes for multiple-antenna systems. It is well known that orthogonal space- time codes have full diversity and linear decoding complexity. They are particularly suitable for transmissions in the network setting using distributed space-time coding since their 'scale-free' property leads to good performance. Our simulations show that they achieve lower error rates than the random code. We also compare the performance of distributed space-time coding to that of selection decode-and-forward using the same orthogonal designs. Simulations show that distributed space-time coding achieves higher diversity than selection decode-and-forward when there are more than one relay node.

We apply the idea of space-time coding devised for multiple antenna systems to the communication over a wireless relay network. We use a two stage protocol, where in one stage the transmitter sends information and in the other, the relay nodes encode their received signals into a "distributed" linear dispersion code, and then transmit the coded signals to the receiver. We show that for high SNR the pairwise error probability (PEP) behaves as (log P/P)/sup min{T,R}/ with T the coherence interval, R the number of relay nodes and P the total transmit power. Thus, apart from the log P factor and assuming T/spl ges/R, the system has the same diversity as a multi-antenna system with R transmit antennas, which is the same as assuming that the R relay nodes can fully cooperate and have full knowledge of the transmitted signal. We further show that for a fixed total transmit power across the entire network, the optimal power allocation is for the transmitter to expend half the power and for the relays to collectively expend the other half. We also show that at low and high SNR, the coding gain is the same as that of a multi-antenna system with R antennas. At intermediate SNR, it can be quite different, which has implications for the design of distributed space-time codes.

In [1], the idea of distributed space-time coding was proposed to achieve a degree of cooperative diversity in a wireless relay network. In particular, for a relay network with a single-antenna transmitter and receiver and R single-antenna relays, it was shown that the pairwise error probability (PEP) decays as ((log P)/P)^R, where P is the total transmit power. In this paper, we extend the results to wireless relay networks where the transmitter, receiver, and/or relays may have multiple antennas. Assuming that the transmitter has M antennas, the receiver has N antennas, the sum of all the antennas at the relay nodes is R, and the coherence interval is long enough, we show that the PEP behaves as (1/P)^(min{M,N}R), if M ≠ N, and ((log^(1/M)P)/p)^(MR), if M=N. Therefore, for the case of M ≠ N, distributed space-time coding has the same PEP performance as a multiple-antenna system with min{M, N}R transmit and a single receive antenna. For the case of M = N, the penalty on the PEP compared to a multiple-antenna system is a log^(1/M) P factor, which is negligible at high SNR. We also show that for a fixed total transmit power across the entire network, the optimal power allocation is for the transmitter to expend half the power and for the relays to share the other half with the power used by each relay being proportional to the number of antennas it has.