While massive multiple-input multiple-output system (MIMO) promises to provide substantial increase in spectral efficiency-per- cell, due to high dimensionality, estimating its channel is considered as one of the major challenges towards extracting all the benefits of such large antenna systems. In this paper, based on parametric channel modeling, we present a direction of arrival (DoA) estimation method for multi-cell multi-user 3D massive-MIMO/Full Dimension (FD-MIMO) orthogonal frequency division multiplexing (OFDM) system using estimation of signal parameter via rotational invariance technique (ESPRIT). Furthermore, we analytically characterize the performance of ESPRIT-based DoA estimation in multiuser scenario, and investigate how pilot contamination, intra-cell, and inter-cell interference affect DoA estimation performance.

Reservoir Computing (RC) is a class of machine learning approaches that is suitable for prediction tasks with low computational complexity. In this paper, an RC-based symbol detection for MIMO- OFDM systems is presented where RC is realized through the echo state network (ESN). Detailed energy-efficiency analysis is conducted to characterize the energy-efficiency of the introduced symbol detector. To be specific, the transmit power, the circuit power, and the computational power at both transmitter and receiver are jointly considered for the energy-efficiency analysis. The overall system energy-efficiency as well as the receiver energy-efficiency of the introduced RC-based symbol detector are compared with those of the popular linear minimum mean squared error (LMMSE)-based approach. Simulation and numerical results show that the RC-based symbol detector is a ``green'' solution compared to the traditional LMMSE-based method with lower energy consumption per information bit.

Mobile health (m-Health) is one potential application of cyber-physical systems, where biomedical sensors and mobile devices interact tightly together to transmit medical data to an m-Health server. In this paper, we consider a three-tier hierarchical m-Health system: 1) the sensor network tier capturing vital signals; 2) the mobile computing network tier processing and routing the sensed data to a fixed remote location; and 3) the back-end network tier processing and analyzing the sensed medical data along with patient's medical history. Based on this architecture, a physical layer security scheme for the second tier is developed and network performance of the introduced scheme is analyzed under different metrics using stochastic geometry. To be specific, secure transmission range and average end-to-end delay are analyzed for two different strategies: the mobile device transmits: 1) to the nearest neighbor and 2) to the furthest neighbor. Furthermore, we consider the cases of full knowledge on eavesdroppers' locations and when such information is unavailable. Results show that transmitting to the nearest neighbor achieves the highest secure transmission distance with the lowest mean delay when full information on eavesdroppers is available.

In this paper, an optimal power and time assignment strategy is developed for hybrid automatic-repeat-request (H-ARQ) communication protocol over quasi-static Rayleigh fading channels. For any given average total time duration and energy budget, we try to find the sequences of power values and time durations for H-ARQ retransmission rounds that minimize the outage probability of the H-ARQ protocol. We solve the joint optimization of power and time assignment for the H-ARQ protocol and derive a set of equations that describe the optimal transmission power and time duration sequence which enable a recursive calculation. Numerical results show that the performance of the proposed optimal-time and optimal-power assignment scheme is substantially better than that of the conventional equal-time and equal-power scheme as well as the equal-time and optimal-power scheme. For example, when the maximum transmission number is L=2 and the target data amount per packet is D_0=2000 bits, the optimal-time and optimal-power assignment scheme saves about 28dB in terms of the average energy consumption.

We propose a novel spectrum decision scheme (i.e., channel selection and handoff) for wireless mesh networks (WMN) which use multiple channels and nodes equipped with multi-beam directional antennas. Our scheme has the following features: (i) It performs spectrum decision by considering various WMN parameters, including the channel quality, beam orientation, antenna-caused deafness and capture effects, and application priority level. (ii) It uses the reinforcement learning (RL)-based spectrum decision process to achieve the optimal quality of multimedia transmission in the long term. However, a newly-joined WMN node could take a long time to make a correct spectrum decision due to the difficult choice of initial RL parameters. Therefore, our scheme uses the apprenticeship learning in conjunction with the RL model, to speed up the spectrum decision process by choosing a suitable neighboring node (called “expert”) to teach a newly-joined node (called “apprentice”). Our experiments demonstrate that the proposed spectrum decision scheme improves the network performance and multimedia transmission quality.

Energy efficiency is one of the major challenges of 5G networks. As data rates are expected to increase by 1000x from 4G to 5G, energy efficiency will need to improve by about the same amount. Recently, energy harvesting techniques have attracted lots of attention from the scientific community, due to their ability to increase network lifetime. More specific, energy harvesting from ambient radio frequency (RF) signals is of special importance, especially with the recent RF circuit advancements. In this paper, we consider a device-to-device (D2D) communication in underlay cellular networks, where D2D users reuse the spectrum occupied by cellular users. We introduce the concept of spatial RF energy harvesting, where D2D users harvest RF power from uplink cellular transmissions, if it exceeds a predesigned threshold, in a spatial region. Using tools from stochastic geometry, we obtain a closed form expression for the probability of activating RF power conversion circuit by making full use of spatial locations of ambient RF signals. Subsequently, we study the impact of RF energy harvesting region radius to harvest sufficient power on the signal-to-interference (SIR) ratio of D2D network. Simulation results provide insights for the required advancements to design highly efficient RF harvesting circuits.

With the explosive growth of mobile traffic demand, device-to-device (D2D) communication offers a promising approach to reduce cellular congestion by offloading users' traffic to D2D network. In this paper, we consider a relay- assisted D2D communication in underlay cellular networks, where D2D users access a proportion of the spectrum occupied by cellular users. This poses two fundamental issues. The first issue is related to protecting the cellular transmissions by reducing the spectrum partition factor, which measures the fraction of spectrum available to D2D users. The second issue is that as we limit the spectrum available to D2D users, less D2D transmissions can be supported, which in turn increases the distances between them. To address these issues, we allow a relay user to assist D2D communications in order to achieve higher D2D spectral efficiency without affecting cellular spectral efficiency. In fact, results show significant improvements in spectral efficiency not only to D2D, but to the whole network, especially in a dense D2D environment.

In this paper, we consider device-to-device (D2D) communication underlaying cellular networks, where D2D users harvest radio frequency (RF) power from uplink cellular transmissions. This paper addresses two important issues for energy harvesting-based D2D cellular networks. The first is how the energy harvested from cellular ambient RF signals affects the D2D spectral efficiency. Using tools from stochastic geometry, we investigate this issue by first characterizing the transmission probability of D2D transmitter as the probability of having enough battery power, and then obtaining analytic expressions of the spectral efficiency for D2D and cellular networks. The second issue is related to the cellular spectral efficiency: the more RF energy D2D users can harvest, the higher their transmission probability becomes, which leads to generating more interference in the cellular network. We carry out simulations to understand this impact of RF energy harvesting on the spectral efficiency of both D2D and cellular networks. Results show promising improvement to the whole network, in terms of the weighted spectral efficiency, when employing RF harvesting technology and when there are enough available channels in the network.

Ultra-densification is one of the main features of 5G networks. In an ultra-dense network, how to conduct interference management and spectrum allocation is a challenging issue. Spectrum sensing in cognitive radio networks is a distributed and efficient way to resolve this issue in ultra-dense networks. However, most of the studies on spectrum sensing only focus on sensing temporal spectrum opportunities where one or multiple primary users are active, which does not make full use of spectrum opportunities in the spatial location domain. To overcome the shortcomings of conventional temporal spectrum sensing, we study the problem of spatial spectrum sensing, which senses spatial spectrum opportunities in wireless networks. In this paper, the performance of spatial spectrum sensing and its application in sensing-based device-to-device (D2D) cellular networks are analyzed using stochastic geometry. Specifically, by modeling the locations of active transmitters as a Poisson point process, the spatial spectrum sensing problem is formulated using the framework of a detection theory. Closed-form expressions are obtained for the sensing threshold, probabilities of spatial detection, and false alarm. Furthermore, analytical throughput for D2D users and cellular users under both channel inversion and constant power allocation cases are derived. The optimal sensing radius that maximizes the defined network metric is obtained numerically. Finally, the simulation and numerical results are presented to verify our theoretical analysis.

To suppress the co-channel interference in a multi-cell multi-user multiple-input-multiple-output downlink cellular network, a novel interference alignment transceiver beam-forming design along with a low complexity iterative coordinated beam-forming scheme is introduced. While the latter combats the intra-cell interference, the former is utilized to mitigate the inter-cell interference. The proposed schemes consider the codebook-based feedback, which is adopted in the LTE/LTE-advanced systems. Optimal downlink user-specific and cell-specific beam-forming matrices are characterized to maximize the lower bound of expected signal-to-leakage-plus-noise ratio and to minimize the residual inter-cell interference, respectively. Moreover, closed-form expressions for these beam-forming matrices under limited channel state information feedback and in the presence of the quantization error are identified. Simulations are conducted to investigate the performance of the proposed strategy. The results indicate that our scheme can significantly improve the average spectral-efficiency of the underlying network when compared with existing ones where the quantization error is neglected. Furthermore, for a fixed payload size of the codebook, unlike zero-forcing beam-forming in which the sum throughput is bounded as the signal-to-noise ratio (SNR) increases, in our scheme, the performance gap between rank 2 feedback and perfect feedback remains approximately constant as SNR increases.

Differential amplify-and-forward (DAF) cooperative relaying appears to be an attractive strategy for wireless networks where channel estimation is not feasible or it is rather avoided. In this work, we intend to determine the optimum relay location in order to minimize the outage probability of the DAF cooperative relaying in case that the system is allowed to deploy the relay. We consider two scenarios. First, we optimize the relay location for the DAF cooperative relaying with any given power assignment at the source and the relay. Second, we further consider relay location optimization with optimum power assignment in which the performance of the DAF relaying is further enhanced with the price of higher complexity. In order to obtain the optimum relay location, we develop a recursive algorithm based on the fix-point theorem and show that the relay location has a unique optimum solution which can be determined recursively from the sequence developed in our theorem. We also study the impact of the path-loss exponent to the optimal relay location and found that the relay should be deployed closer to the source in lossy environment such as near-ground environment than that in the free space environment. Extensive numerical results are provided to validate and illustrate the theoretical development.

We present a real-time, priority-aware joint packet fragmentation and error protection scheme for transmitting H.264/AVC compressed video over Rayleigh fading channels. A priority is assigned to every video slice based on the predicted cumulative mean squared error (CMSE) contributed by its loss relative to all the slices of the group of pictures (GOP). Here, the CMSE of a slice, which is predicted by using our low-complexity generalized linear model, represents the degradation in video quality contributed by its loss. The slices of a frame belonging to the same priority are aggregated to form video packet(s). We simulate the fragment error rates (FERs) for a combination of different fragment sizes and rate compatible punctured convolutional (RCPC) code rates. These FERs are then used to determine the optimal fragment sizes and code rates for packets of each priority class by minimizing the expected normalized predicted CMSE of all the priority classes per GOP in H.264 video bitstream. We observed a significant improvement in the received video quality over the conventional and priority-agnostic packet fragmentation schemes.

We formulate the problem of point-to-point scheduling at a congested network node as a Markov decision process (MDP) that considers the deadlines and priorities of each packet as well as the dynamic packet arrivals and channel conditions. Within this framework, we formulate the problem with the objective of maximizing the node's long-run priority-weighted throughput subject to instantaneous transmission rate constraints. We then analyze the structural properties of the optimal scheduling policy with respect to the deadlines and priorities of the backlogged packets. Additionally, we compare our approach to existing heuristics such as Priority Queueing (PQ), Earliest Deadline First (EDF), and Weighted Fair Queueing (WFQ). Our MDP-based approach outperforms all three heuristics not only because it takes into account the packets' priorities and deadlines, but also because it takes into account the future channel and packet arrival dynamics. Lastly, we experimentally show that the optimal scheduling policy has a switch-over type structure in several key parameters including the relative priorities of different traffic classes, the discount factor, and the traffic load intensity.

In underlay cognitive radio networks (CRNs), secondary users (SUs) have to dynamically control their transmit powers so that the interference to primary users (PUs) is tolerable. Under this constraint, SUs' data link usually suffers from either high error rate with limited transmission range or long end-to-end delay caused by multi-hop transmissions. To improve the performance of SUs, cooperative routing using mutual-information accumulation is introduced in CRNs for the first time in this paper. To be specific, the routing and resource allocation problem in underlay CRNs is investigated and is factored into two sub-problems, each of which can be solved efficiently. Furthermore, a distributed algorithm is introduced and simulation results show that both the centralized and distributed algorithms can reduce upto 77% of the end-to-end delay compared to the traditional multi-hop delay-optimal routing in CRNs. Finally, theoretical analysis on the end-to-end delay for a one-dimensional (1-D) network in the low signal-to-interference ratio (SIR) region is conducted. Both the analytical and simulation results show that mutual-information accumulation can significantly decrease the delay of underlay CRNs, especially in the scenario where PUs have a tight interference power constraint.

To improve the spectral efficiency of a cognitive radio system, sensing based spectrum sharing (SBSS) technique combines the advantages of spectrum overlay and spectrum underlay. In this paper, we study the performance of SBSS under primary users' (PUs') rate loss constraint. To be specific, efficient algorithms are introduced to find the optimal sensing time and power allocation in both single-carrier and multi-carrier systems. Performance evaluation is conducted to compare the system throughput between SBSS and spectrum overlay techniques. Simulation results suggest that SBSS outperforms spectrum overlay when the transmitter of the secondary user (SUT) is far away from the receiver of the primary user (PUR) in the single-carrier case. In the multi-carrier case, SBSS can achieve 15.2% increase of spectral efficiency over spectrum overlay when all the channel gains are independent and identically distributed.

Cooperative communication has emerged as a new wireless network communication concept, in which parameter optimization such as power budget and time allocation plays an important role in cooperative relaying protocol designs. While most existing works on cooperative relaying protocol designs considered equal-time allocation scenario, i.e. equal time duration is assigned to each source and each relay, in this work we intend to design and optimize cooperative communication protocols by exploring all possible variations in time and power domains. We jointly optimize the power and time allocations for the cooperative relaying protocol such that the outage probability of the protocol is minimized. Specifically, for any given time allocation, we are able to determine the corresponding optimum power allocation analytically with a closed-form expression. We also show that in order to minimize the outage probability of the protocol, one should always allocate more energy and time to the source than the relay. Extensive numerical and simulation studies illustrate our theoretical developments.

Cooperative communication has emerged as a new wireless network communication concept, in which parameter optimization such as power budget and time allocation plays an important role in cooperative relaying protocol designs. While most existing works on cooperative relaying protocol designs considered equal-time allocation scenario, i.e. equal time duration is assigned to each source and each relay, we intend to design and optimize cooperative communication protocols by exploring all possible variations in time and power domains. In our recent work [18], we explored an ideal non-equal-time cooperative relaying protocol where the system can use arbitrary re-encoding methods at the relay and adjust time allocation arbitrarily, in which we showed that with optimum power and time allocation the cooperative relaying protocol has significant performance improvement compared to the equal-time relaying protocol. In this paper, with more realistic consideration, we design a practical cooperative relaying protocol based on linear mapping, i.e. using linear mapping as the re-encoding method at the relay and considering integer time slots in the two phases. Furthermore, we develop an optimum linear mapping to minimize the outage probability of the linear-mapping based cooperative relaying protocol. Numerical and simulation results show that the performance of the proposed cooperative relaying protocol based on the optimum linear mapping is close to the performance benchmark of the ideal cooperative protocol.

In this paper, we systematically analyze the outage probability performance of the distributed cooperative multicast with randomized orthogonal space-time codes (OSTC) for wireless multimedia networks. The distributed cooperative multicast protocol with randomized OSTC enables users who successfully decode signals from base station or access point serving as relays to forward the signals to other users, and consequently the overall performance of wireless multimedia multicast can be greatly improved. We derive a closed-form expression for the outage probability for the distributed cooperative multicast with randomized OSTC. To get more insight, we further develop an asymptotically tight approximation for the outage probability which reveals the diversity order of the protocol as K + 1, where K is the size of the randomized OSTC used in the protocol. Based on the asymptotically tight approximation of the outage probability, we are able to determine an optimum power allocation for the distributed cooperative multicast protocol. It shows that in a statistic sense, we should allocate 1/(K + 1) of the total power budget to the BS/AP and evenly distribute the rest for the users as relaying power. Numerical and simulation results validate our theoretical development.

Enabling the spectrum handoff for multimedia applications in cognitive radio networks (CRNs) is challenging, due to multiple interruptions from primary users (PUs), contentions among secondary users (SUs), and heterogenous Quality-of-Experience (QoE) requirements. In this paper, we propose a learning-based and QoE-driven spectrum handoff scheme to maximize the multimedia users' satisfaction. We develop a mixed preemptive and non-preemptive resume priority (PRP/NPRP) M/G/1 queueing model for modeling the spectrum usage behavior for prioritized multimedia applications. Then, a mathematical framework is formulated to analyze the performance of SUs. We apply the reinforcement learning to our QoE-driven spectrum handoff scheme to maximize the quality of video transmissions in the long term. The proposed learning scheme is asymptotically optimal, model-free, and can adaptively perform spectrum handoff for the changing channel conditions and traffic load. Experimental results demonstrate the effectiveness of the proposed queueing model for prioritized traffic in CRNs, and show that the proposed learning-based QoE-driven spectrum handoff scheme improves quality of video transmissions.

The unequal error protection (UEP) has shown promising results for transmitting video over error-prone wireless channels. In this paper, we investigate the cross-layer design of forward error correction (FEC) schemes by using the UEP Raptor codes at the application layer (AL) and UEP rate compatible punctured convolutional (RCPC) codes at physical layer (PHY) for prioritized video packets. The video packets are prioritized based on their contribution to the received video quality. A genetic algorithm (GA)-based optimization algorithm is proposed to find the optimal parameters for both Raptor and RCPC codes, to minimize the video distortion and maximize the peak signal-to-noise-ratio for the given video bit rates and channel constraints (i.e., SNR and available bandwidth). We evaluate the performance of four combinations of the UEP schemes for H.264/AVC encoded video sequences over the AWGN and Rayleigh fading channels and show the superiority of the optimized cross-layer UEP FEC scheme. For Rayleigh fading channel, the proposed cross-layer optimization uses two different time-scales at AL and PHY which allows PHY to adapt faster to the changing channel quality.

Cooperative communication has emerged as a new wireless network communication concept, in which parameter optimization such as power budget and time allocation plays an important role in cooperative relaying protocol designs. While most existing works on cooperative relaying protocol designs considered equal-time allocation scenario, i.e., equal time duration is assigned to each source and each relay, in this work we intend to design and optimize cooperative communication protocols by exploring all possible variations in time and power domains. We consider a cooperative relaying network in which no channel state information (CSI) is available at the transmitter side and the protocol optimization is based on channel statistics (i.e., mean and variance) and it does not depend on instantaneous channel information. First, we consider an ideal cooperative relaying protocol where the system can use arbitrary re-encoding methods at the relay and adjust time allocation arbitrarily. We obtain an optimum strategy of power and time allocations to minimize the outage probability of the ideal cooperative protocol. Specifically, for any given time allocation, we are able to determine the corresponding optimum power allocation analytically with a closed-form expression. We also show that to minimize the outage probability of the protocol, one should always allocate more energy and time to the source than the relay. Second, with more realistic consideration, we design a practical cooperative relaying protocol based on linear mapping, i.e., using linear mapping as the re-encoding method at the relay and considering integer time slots in the two phases. The theoretical results from the ideal cooperative protocol serve as a guideline and benchmark in the practical cooperative protocol design. We also develop an optimum linear mapping to minimize the outage probability of the linear-mapping based cooperative protocol. Extensive numerical and simulation studies illustrate our theoretical developments and show that the performance of the proposed cooperative relaying protocol based on the optimum linear mapping is close to the performance benchmark of the ideal cooperative protocol.

In this paper we consider the problem of resource management for a Direct Sequence Code Division Multiple Access (DS-CDMA) wireless Visual Sensor Network (VSN) in a generalized fading environment. In a VSN application, the primary goal is ensuring that maximum video quality is achieved in spite of the prevailing network resource constraints. The Nash Bargaining Solution (NBS) was used in determining the transmission power and source and channel coding rates for each node. The nodes in the network negotiate in order to determine their transmission parameters. The task is to optimize the transmission powers (which are continuous) and the source and channel coding rates (which are discrete) for all the network nodes. Particle Swarm Optimization (PSO) is used to solve the mixed-integer optimization that arises. The analysis was carried out for a myriad of wireless multipath fading environments using a unified moment generating function (MGF) approach.

Cognitive Radio (CR) can efficiently utilize the licensed wideband spectrum whenever the primary users (PUs) are absent. Spectrum sensing is the first step and an important function to fulfill the CR system. A cyclostationary feature detector can robustly detect the PU's modulated signals even under strong interferences. However this detector needs high signal sampling rate and also puts heavy computation burden on the system. Compressive sensing (CS) can compress the data at the front sampling end but has high overload and delay from the reconstruction side. In this work we generate the compressive CR spectrum measurement by utilizing both the cyclostationary feature and sparsity prior knowledge at the spectrum sensing front end, and we apply the compressive signal processing (CSP) without the need of signal or feature reconstruction. This can significantly shorten the CR spectrum sensing time. Our experimental results have shown the pattern analysis accuracy and efficiency of our CSP scheme.

This paper investigates the linear precoder design for multiple-input multiple-output (MIMO) K-user interference channels with finite alphabet inputs. We first obtain the general explicit expressions of the achievable rate of each user in MIMO interference channel systems. We study optimal transmission strategies in both high signal-to-noise ratio (SNR) and low SNR regions. We show that given finite alphabet inputs, a simple power allocation design can achieve optimal performance. In contrast, the well-known interference alignment technique for Gaussian input scenarios, only utilizes a partial interference-free signal space for transmission and leads to a constant performance loss when it is applied to finite-alphabet input scenarios. We determine this constant rate loss at high SNR. Moreover, we establish necessary conditions for the linear precoder design of the weighted sum-rate maximization. We also develop an efficient iterative algorithm for determining precoding matrices of all the users. Our numerical results show that for the practical digital modulated signals from discrete constellations, the proposed iterative algorithm achieves considerably higher sum-rate than the existing methods.

The problem of resource management in a Direct Sequence Code Division Multiple Access (DS-CDMA) wireless Visual Sensor Network (VSN) with M-array Phase Shift Keying (MPSK) modulation in an Additive White Gaussian Network (AWGN) channel was considered in this paper. Achieving maximum video quality, in spite of the prevailing network resource constraints, is of utmost importance in VSN applications. Our optimization scheme is based on the Nash Bargaining Solution (NBS). The nodes in the network negotiate in order to determine their transmission parameters (transmission powers; source and channel coding rates for each node). The task is to optimize the transmission powers (which are continuous) and the source and channel coding rates (which are discrete) for all the network nodes, while taking advantage of the improved bandwidth spectral efficiency provided by the higher order constellation.

This paper investigates the linear precoder design for K-user interference channels of multiple-input multiple-output (MIMO) transceivers under finite alphabet inputs. We first obtain general explicit expressions of the achievable rate for users in the MIMO interference channel systems. We study optimal transmission strategies in both low and high signal-to-noise ratio (SNR) regions. Given finite alphabet inputs, we show that a simple power allocation design achieves optimal performance at high SNR whereas the well-known interference alignment technique for Gaussian inputs only utilizes a partial interference-free signal space for transmission and leads to a constant rate loss when applied naively to finite-alphabet inputs. Moreover, we establish necessary conditions for the linear precoder design to achieve weighted sum-rate maximization. We also present an efficient iterative algorithm for determining precoding matrices of all the users. Our numerical results demonstrate that the proposed iterative algorithm achieves considerably higher sum-rate under practical QAM inputs than other known methods.

We consider a secondary multi-hop cognitive radio network with decentralized control that operates cognitively to coexist with primary users. We propose a new spread-spectrum management paradigm, in which, unlike mainstream dynamic spectrum access research, digital waveforms are designed to occupy the entire available spectrum, and to adaptively track the interference profile at the receiver to maximize the link capacity while avoiding interference to primary users. In this context, we study the problem of maximizing the network throughput of a multi-hop network through joint routing and spread-spectrum channelization. We first propose a centralized formulation of the network control problem. We then propose an algorithm that can be seen as a distributed localized approximation of the throughput-maximizing policy. We refer to the proposed jointly-designed routing and code-division channelization algorithm as ROCH (Routing and cOde-division CHannelization). Specifically, power and spreading code are jointly selected to maximize the pre-detection secondary \mathrm{SINR} while providing quality of service guarantees to on-going primary and secondary transmissions, while the routing algorithm dynamically selects relays based on the network traffic dynamics and on the achievable data rates on different secondary links. We study the throughput and delay performance of ROCH through a extensive simulation experiments, which demonstrate the appeal of the proposed framework through significant performance gains compared to baseline solutions.

In this paper, the capacity of airborne multiple-input-multiple-output (MIMO) wireless communication systems with arbitrary alignments of linear transmit and receive antenna arrays is systematically analyzed and the maximum achievable capacity is determined. Based on a general three-dimensional (3D) airborne MIMO communication model, we are able to approximate the airborne MIMO capacity as a function of the transmit and receive antenna array geometry in the 3D space. The capacity approximation is asymptotically tight as the distance between the transmit and receive antenna arrays large compared to their size. Based on the asymptotically tight capacity approximation, we derive an upper bound as well as a lower bound of the airborne MIMO capacity. Interestingly, both the upper and lower bounds are achievable. We also derive a necessary and sufficient condition for airborne MIMO communication systems to achieve the capacity upper bound for any given 3D transceiver antenna array geometry. The necessary and sufficient condition allows us to properly select the system parameters and design airborne MIMO communication systems that reach the best possible performance in terms of system capacity. We prove that when the distance between the transmit and receive antenna arrays is within a certain range, there exists a set of system parameter values (e.g. antenna element separation) for which the capacity of the MIMO communication system achieves the theoretical upper bound and this capacity value is larger than the average capacity of the corresponding conventional MIMO communication system under Rayleigh fading. Finally, we prove that the airborne MIMO capacity converges to the capacity lower bound when the distance between the transmit and receive antenna arrays goes to infinity. Extensive numerical studies included in this paper illustrate and validate our theoretical developments.

In this paper, a penalty-based distributed algorithm is proposed to design beamforming for multi-source multi-destination (MSMD) networks. The objective of the beamforming design is to minimize the transmission power of each source under the constraint that the signal-to-interference-plus-noise ratio (SINR) requirements of all source-destination pairs are satisfied. The proposed beamforming design has better power-efficiency compared to the egoism, altruism, and balancing algorithms. The convergence of the proposed algorithm is also discussed and a sufficient condition is determined for the algorithm to converge to a Pareto optimal solution. Simulation results show that the proposed algorithm has higher probability of convergence compared to the egoism, altruism, and balancing algorithms, which is close to a centralized performance benchmark.

In this paper, we develop a methodology to analyze systematically the capacity of airborne MIMO wireless communication systems with linear transmit and receive antenna arrays. Our methodology is based on a general three-dimensional airborne MIMO communication model that we develop to accommodate arbitrary alignments of the transmit and receive antenna arrays. Specifically, we determine an upper bound and a lower bound for the airborne MIMO capacity. Then, we derive a necessary and sufficient condition for airborne MIMO communication systems to achieve the capacity upper bound. The necessary and sufficient condition allows us to properly select the system parameters and design airborne MIMO communication systems that reach the best possible performance in terms of system capacity. We prove that when the distance between the transmit and receive antenna arrays is within a certain range, there exists a set of system parameter values for which the capacity of the MIMO communication system achieves the theoretical upper bound and this capacity value is larger than the average capacity of the corresponding conventional MIMO communication system with Rayleigh fading. We also show that asymptotically, as the distance between the transmit and receive antenna arrays goes to infinity, the capacity of the airborne MIMO system converges to the capacity lower bound. Extensive numerical studies included in this paper illustrate and validate our theoretical developments.

In this paper, the optimal power assignment strategy is determined for hybrid automatic-repeat-request (H-ARQ) protocols such that the average delay of the protocol is minimized for any given total transmission power budget and any targeted outage probability. A set of equations is derived that describe the optimal transmission power sequence and its optimality is shown based on the Karush-Kuhn-Tucker (KKT) Theorem. The set of equations enables an exact recursive calculation of the optimal transmission power per round, and the calculation complexity is fixed regardless of the maximum number of (re)transmission rounds allowed in the H-ARQ protocol. Compared to the conventional equal power assignment strategy, the optimal power assignment scheme achieves the same average delay with much less total transmission power. More importantly, for certain power budget levels, the optimal power assignment can make the H-ARQ protocol work while the equal power assignment cannot. Extensive numerical results are presented to illustrate the theoretical development.

This paper considers the problem of blindly extracting data embedded over a wide band in a spectrum (transform) domain of a digital medium (image, audio, video). We first develop a multi-signature iterative generalized least-squares (M-IGLS) core procedure to seek unknown data hidden in hosts via multi-signature direct-sequence spread-spectrum embedding. Neither the original host nor the embedding signatures are assumed available. Then, cross-correlation enhanced M-IGLS (CC-M-IGLS), a procedure described herein in detail that is based on statistical analysis of repeated independent M-IGLS processing of the host, is seen to offer most effective hidden message recovery. Experimental studies on images show that the proposed CC-M-IGLS algorithm can achieve recovery probability of error close to what may be attained with known embedding signatures and host autocorrelation matrix.

The quality of H.264/AVC compressed video delivery over wireless channels is affected by packet losses. Aggregating H.264/AVC slices to form video packets with sizes adaptive to their importance can improve transmission reliability. Larger packets are more likely to be in error but smaller packets cause more overhead. A second method is assigning stronger channel code rates to more important slices. We use cross-layer dynamic programming to address both adaptive packet formation as well as RCPC-channel code rate allocation simultaneously, to improve received video quality. Simulation results show the advantages of the proposed scheme.

We propose a low complexity Generalized Linear Model (GLM) for prioritizing slices during real-time H.264/AVC compressed video streaming. We train the GLM over a video database to predict the Cumulative Mean Square Error (CMSE) corresponding to individual slice losses by using a combination of efficient video parameters which can be easily extracted during the encoding of a frame. We prioritize the slices generated within a Group of Pictures (GOP) based on the predicted CMSE by using a Quartile Based Prioritization (QBP) scheme. For comparison, we also perform QBP on measured CMSE values from individual pre-encoded slice losses and analyze the priority misclassifications of the slices. We validate our model by applying Unequal Error Protection (UEP) using RCPC codes to the different prioritized bitstreams and evaluating their performance over noisy channels. Simulation results show that predicted CMSE schemes achieve PSNR performance close to that of the measured CMSE schemes for different slice sizes, video bitrates and over different channel SNRs.

In this paper, we consider the problem of designing transceiver beamforming vectors for multi-source multi-destination (MSMD) wireless networks such that the transmission power of each source is minimized while the signal-to-interference plus noise ratio (SINR) requirements of all source-destination pairs are satisfied. We propose an efficient iterative algorithm to design the transceiver beamforming vectors and address the convergence of the algorithm. We determine a necessary condition as well as a sufficient condition for the algorithm to converge to a generalized Nash equilibrium solution. Especially, if each destination has only one antenna, we obtain a necessary and sufficient condition for the algorithm to converge to a unique generalized Nash equilibrium solution. Simulation results show that the proposed iterative algorithm has higher probability of convergence compared to an iterative waterfilling (IWF) approach. For example, for a system with three source-destination pairs and SINR requirement of 2dB, the probability of convergence is 84% with the proposed algorithm while it is only 66% with the IWF algorithm.

In this work, we consider the problem of assigning optimal transmission power sequence for cooperative hybrid automatic-repeat-request (H-ARQ) relaying protocol over quasi-static Rayleigh fading channels. We try to determine the optimal power sequence by minimizing the average total transmission power that we analyzed in our previous work. However, the closed-form expression of the average total power consumption of the cooperative H-ARQ relaying protocol is complicated in general, so we develop first in this work a simple approximation of the average total transmission power that is tight at high SNR. Then, based on the asymptotically tight approximation, we are able to identify the sequence of power values that minimizes the average total power consumption of the cooperative H-ARQ relaying protocol for any given targeted outage probability. In particular, we derive a set of equations that describe the optimal power level in each (re)transmission and enable its recursive calculation with fixed searching complexity. When the maximum number of (re)transmissions allowed in the protocol is L = 2, we have a closed-form result for the optimal transmission power sequence. The optimal power assignment solution reveals that conventional equal power assignment scheme is not optimal in general. Extensive simulation and numerical results are provided to illustrate and validate the theoretical results.

In this paper, we study the problem of finding the maximum throughput achievable by secondary users while primary users are guaranteed to achieve the minimum required data rates. We assume that node positions and channel side informations are available, and a virtual central unit is in responsible for scheduling both primary and secondary users. The optimal solution provides a theoretical upper bound of throughput for cognitive radio networks. An iterative algorithm proposed in [1] was extended to solve the throughput optimization problem, which can be decomposed into sub-problems and solved iteratively. However, due to the NP-hardness of Maximum Weighted Independent Set (MWIS) problem which has to be solved at each iteration, some greedy approximation method with polynomial computation time is applied in the iterative algorithm. Finally, we derive a theoretical upper bound for the distance between the greedy solution and the optimal solution, and guarantee that the optimal solution must be within this upper bound of the greedy solution. Numerical results are given to validate the solution.

The capacity optimization problem of MIMO links with interference has attracted an increasing interest. Due to the nonconvexity of the capacity problem, only suboptimal solutions can be found. In the previous works, a Gradient Projection (GP) algorithm [1] and a Quasi-Newton (QN) method [2] were proposed to provide suboptimal solutions subject to the constant power constraint. In this paper, we derive the capacity for MIMO links decomposed via SVD and interfered from other links. Then, each eigenchannel of MIMO link is represented by a set of logical links with a set of discrete data rates and discrete powers. An Integer Programming based algorithm (named as IP) is presented to solve the capacity optimization problem. The solution specifies the set of logical links that can transmit simultaneously. Numerical results show that GP and QN methods achieve better performance than IP method for the case of weak interference because of the convexity of the optimization problem when INR is sufficiently small. In the case of strong interference, IP method achieves better performance than GP and QN methods, which means that transmitting one link at a time is better than transmitting all links simultaneously with full power. In other words, scheduling links to transmit is more important for the case of strong interference.

We address the fundamental problem of identifying the optimal power assignment sequence for hybrid automatic-repeat-request (H-ARQ) communications over quasi-static Rayleigh fading channels. For any targeted H-ARQ link outage probability, we find the sequence of power values that minimizes the average total expended transmission power. We first derive a set of equations that describe the optimal transmission power assignment and enable its exact recursive calculation. To reduce calculation complexity, we also develop an approximation to the optimal power sequence that is close to the numerically calculated exact result. The newly founded power allocation solution reveals that conventional equal-power H-ARQ assignment is far from optimal. For example, for targeted outage probability of 10-3with a maximum of two transmissions, the average total transmission power with the optimal assignment is 9 dB lower than the equal-power protocol. The difference in average total power cost grows further when the number of allowable retransmissions increases (for example, 11 dB gain with a cap of 5 transmissions) or the targeted outage probability decreases (27 dB gain with outage probability 10-5and transmissions capped at 5). Interestingly, the optimal transmission power assignment sequence is neither increasing nor decreasing; its form depends on given total power budget and targeted outage performance levels. Extensive numerical and simulation results are presented to illustrate the theoretical development.

In this paper, we propose an approach to manage network resources for a direct sequence code division multiple access (DS-CDMA) visual sensor network where nodes monitor scenes with varying levels of motion. It uses cross-layer optimization across the physical layer, the link layer, and the application layer. Our technique simultaneously assigns a source coding rate, a channel coding rate, and a power level to all nodes in the network based on one of two criteria that maximize the quality of video of the entire network as a whole, subject to a constraint on the total chip rate. One criterion results in the minimal average end-to-end distortion amongst all nodes, while the other criterion minimizes the maximum distortion of the network. Our experimental results demonstrate the effectiveness of the cross-layer optimization.

We consider the problem of simultaneous power and code-channel allocation for a secondary transmitter/receiver pair coexisting with a primary code-division multiple-access (CDMA) system. Our objective is to find the optimum transmitting power and code sequence of the secondary channel that maximize the signal-to-interference-plus-noise ratio (SINR) at the output of the maximum SINR linear receiver, while at the same time the SINR of all primary channels at the output of their max-SINR receiver is maintained above a certain threshold. This is a non-convex NP-hard optimization problem. We propose a novel feasible suboptimum solution using semidefinite programming. Simulation studies illustrate the theoretical developments.

We consider the problem of cognitive code-division channelization (simultaneous power and code-channel allocation) for secondary transmission links co-existing with an unknown primary code-division multiple-access (CDMA) system. We first develop a blind primary-user identification scheme to detect the binary code sequences (signatures) utilized by primary users. To create a secondary link we propose two alternative procedures -one of moderate and one of low computational complexity- that optimize the secondary transmitting power and binary code-channel assignment in accordance with the detected primary code channels to avoid "harmful" interference. At the same time, the optimization procedures guarantee that the signal-to-interference-plus-noise ratio (SINR) at the output of the maximum SINR linear secondary receiver is no less than a certain threshold to meet secondary transmission quality of service (QoS) requirements. The extension of the channelization problem to multiple secondary links is also investigated. Simulation studies presented herein illustrate the theoretical developments.

The current and future wireless systems need to support a multitude of services with a wide range of data rates and reliability requirements. The limited battery resource at the mobile terminals coupled with the hostile multipath fading channel makes the problem of providing reliable multimedia services challenging. In this article, we develop a novel method to mechanize a prioritized unicast transmission at the physical layer in response to the disparate quality of service requirements for multimedia traffic (imposed by upper layers) and investigate how the unequal bit error protection offered by multiresolution modulation is capitalized at the data link layer to yield a substantial throughput gain and minimize the packet loss probability of the "more important" packets in an integrated voice/data and delay-sensitive applications (especially in poor and moderate channel conditions). However, there exists a slight throughput penalty in the high signal-to-noise ratio regime due to competitive nature of the "optimal reliable signaling rate" in an unknown time-varying channel.

In this paper, the total transmission power of a multi-source multi-destination relay network is minimized under the constraint that the signal to interference plus noise ratio (SINR) requirement of each source-destination pair is satisfied. The optimization problem involves K power variables, where K is the number of source-destination pairs in the network, and an exhaustive search is prohibitive for large K. In this work, we develop an asymptotically tight approximation of the SINR that allows us to reformulate the original optimization problem to a single-variable optimization problem, which can be easily solved by numerical search of the single variable. Then, the corresponding optimal transmission power at each source and relay can be calculated directly. The proposed optimization scheme is scalable and leads to a power assignment algorithm that exhibits the same optimization complexity for any number (K) of source-destination pairs in the network. Moreover, for the special case of transmission over orthogonal channels, we derive analytically the solution to the optimization problem. Extensive numerical studies illustrate our theoretical developments.

In this paper, the average total power consumption per information packet is investigated for a cooperative hybrid automatic-repeat-request (H-ARQ) protocol in a quasistatic Rayleigh fading environment. Specifically, a closed-form expression of the average total transmission power is obtained for the cooperative H-ARQ relay protocol, in which the source may use different transmission power level in different (re-)transmission rounds. The closed-form expression is valid for any maximum number of (re-)transmission rounds L allowed by the protocol and may play a key role thereafter in optimizing power allocation. Since the closed-form expression is complicated for large L, an approximation of the average total transmission power is developed which is asymptotically tight at high SNR. Extensive simulation and numerical results are also provided herein to illustrate and validate the theoretical results.

In this paper, a new analytical approach is developed for the evaluation of the outage probability of decode-and-forward (DF) automatic-repeat-request (ARQ) relaying under packet-rate fading (fast fading) channels. Based on this approach, a closed-form asymptotically tight (as SNR → ∞) approximation of the outage probability is derived, and the diversity order of the DF cooperative ARQ relay scheme is shown to be equal to 2L - 1, where L is the maximum number of ARQ retransmissions. The closed-form expression clearly shows that the achieved diversity is partially due to the DF cooperative relaying and partially due to the fast fading nature of the channels (temporal diversity). Numerical and simulation studies illustrate the theoretical developments.

In this work, a cognitive cooperative communication protocol is proposed for cognitive ad-hoc networks, in which primary users and cognitive radio (CR) users may cooperate for mutual benefit. The new cooperation protocol allows active cooperation between primary users and CR users in which CR users assist to relay primary users' signals in exchange for some spectrum released from the primary users. While conventional cognitive radios do not guarantee continuous operation of CR users (they use the spectrum only when primary users do not), the protocol proposed in this work provides continuous service for CR users. The proposed cognitive cooperation protocol is optimized in terms of maximizing the primary user's energy savings and the CR user's own data transmission rate. It turns out that the primary users have significantly average energy savings from cooperation (e.g. up to 50% when compared to a non-cooperation case at the same transmission power level), which provides a good incentive for they to cooperate.

Compressed video data is very sensitive to channel-induced errors and network losses. Most conventional unequal error protection techniques involve forward error correction codes of different rates or asymmetric signal constellations for data of different priorities. It is yet unknown if combining the above two techniques can further improve performance. In this paper, we investigate rate-compatible punctured convolutional (RCPC) codes concatenated with hierarchical QAM for H.264 encoded video sequences. We investigate system constraints and propose an optimization formulation to compute the optimal parameters of the proposed system under the given source significance information. An upper bound to the bit error rate of the proposed system is derived as a function of system parameters, including the code rate and geometry of the constellation. The example shown demonstrates system design for H.264 video, where PSNR improvement is observed.

Differential Amplify-and-Forward (DAF) relaying can be viewed as an attractive cooperative communication strategy for wireless networks where channel estimation is not feasible or it is rather avoided. A new exact outage probability expression is presented herein for DAF relaying that involves only a single integral. Then, an asymptotically tight closed-form approximation is obtained which enables power allocation optimization for the outage behavior of DAF relaying. Numerical and simulation studies validate the theoretical analysis.

We address the fundamental problem of identifying the optimal power allocation sequence for hybrid automatic-repeat-request (H-ARQ) communications over quasistatic Rayleigh fading channels. For any targeted H-ARQ link outage probability, we find the sequence of power values that minimizes the average total expended transmission power. The newly founded power allocation solution reveals that conventional equal-power H-ARQ assignment is far from optimal. For example, for targeted outage probability of 10-3with a maximum of two transmissions, the average total transmission power with optimal assignment is 9dB lower than the equal-power protocol. The difference in average total power cost grows further when the number of allowable retransmissions increases (for example, lldB gain with a cap of 5 transmissions) or the targeted outage probability decreases (27dB gain with outage probability 10-5and transmissions capped at 5).

Tight bounds for the Shannon capacity of amplify-and-forward cooperative diversity networks are derived for three different adaptive source transmission policies in a myriad of fading environments: (i) constant power with optimal rate adaptation (ORA); (ii) optimal joint power and rate adaptation (OPRA); and (iii) fixed rate with truncated channel inversion (TCIFR). Our unified framework based on the moment-generating function (MGF) approach allows us to gain insights as to how fade distributions and dissimilar fading statistics across the distinct communication links will affect the Shannon capacity, without imposing any restrictions on the fading parameters.

Throughput maximization is one of the main challenges in cognitive radio ad hoc networks, where the availability of local spectrum resources may change from time to time and hop-by-hop. Cooperative transmission exploits spatial diversity without multiple antennas at each node to increase capacity with reliability guarantees. This idea is particularly attractive in wireless environments due to the diverse channel quality and the limited energy and bandwidth resources. With cooperation, source node and relay node cooperatively transmit data to the destination. In such a virtual multiple antenna transmission system, the capacity of the cooperative link is much larger than that of the direct link from source to destination. In this paper, we will study decentralized and localized algorithms for joint dynamic routing, relay assignment, and spectrum allocation under a distributed and dynamic environment.

A cognitive radio network with decentralized control (i.e., a cognitive ad hoc network) is considered in this demonstration. The demo implements a decentralized and localized algorithm for through put maximization through joint routing and interference-avoiding waveform selection. The algorithm adapts to time-varying traffic demands, interference profile, and network topology to locally maximize the achievable data rate while avoiding harmful interference to co-located primary or secondary users. The prototype is based on a cross-layer protocol stack implemented in Python, which leverages GNU Radio for adaptive signal generation on a USRP2 software-defined-radio platform.

In this paper we develop a new analytical methodology for the evaluation of the outage probability of cooperative decode-and-forward (DF) automatic-repeat-request (ARQ) relaying under packet-rate fading (fast fading or block fading) channels, where the channels remain fixed within each ARQ transmission round, but change independently from one round to another. We consider a single relay forwarding Alamouti-based retransmission signals in the cooperative ARQ scheme. In particular, (i) we derive a closed-form asymptotically tight (as SNR → ∞) approximation of the outage probability; (ii) we show that the diversity order of the DF cooperative ARQ relay scheme is equal to 2L-1, where L is the maximum number of ARQ (re)transmissions; and (iii) we develop the optimum power allocation for the DF cooperative ARQ relay scheme. The closed-form expression clearly shows that the achieved diversity is partially due to the DF cooperative relaying and partially due to the fast fading nature of the channels (temporal diversity). With respect to power allocation, it turns out that the proposed optimum allocation scheme depends only on the link quality of the channels related to the relay, and compared to the equal power allocation scheme it leads to SNR performance gains of more than 1 dB. Numerical and simulation studies illustrate the theoretical developments.

A new exponential-type integral for the generalized M-th order Marcum Q-function QM(alpha, beta) is obtained when M is not necessarily an integer. This new representation includes a classical formula due to Helstrom for the special case of positive integer order M and an additional integral correction term that vanishes when M assumes an integer value. The new form has both computational utility (numerous existing computational algorithms for QM(alpha, beta) are limited to integer M) and analytical utility (e.g., performance evaluation of selection diversity receiver in correlated Nakagami-m fading with arbitrary fading severity index, unified analysis of binary and quaternary modulations over generalized fading channels, and development of a Markovian threshold model for block errors in correlated Nakagami-m fading channels). Tight upper and lower bounds for QM(alpha, beta) that holds for any arbitrary real order M ges 0.5 are also derived.

The current and future wireless systems need to support a multitude of services with a wide range of data rates and reliability requirements. The limited battery resource at the mobile terminals coupled with the hostile multipath fading channel makes the problem of providing reliable multimedia services challenging. In this article, we develop a novel method to mechanize a prioritized unicast transmission at the physical layer in response to the disparate quality of service requirements for multimedia traffic (imposed by upper layers) and investigate how the unequal bit error protection offered by multiresolution modulation is capitalized at the data link layer to yield a substantial throughput gain and minimize the packet loss probability of the "more important" packets in an integrated voice/data and delay-sensitive applications (especially in poor and moderate channel conditions). However, there exists a slight throughput penalty in the high signal-to-noise ratio regime due to competitive nature of the "optimal reliable signaling rate " in an unknown time-varying channel.

We derive new bounds on the total squared correlation (TSC) of quaternary (quadriphase) signature/sequence sets for all lengths L and set sizes K. Then, for all K, L, we design minimum-TSC optimal sets that meet the new bounds with equality. Direct numerical comparison with the TSC value of the recently obtained optimal binary sets shows under what K, L realizations gains are materialized by moving from the binary to the quaternary code-division multiplexing alphabet. On the other hand, comparison with the Welch TSC value for real/complexfield sets shows that, arguably, not much is to be gained by raising the alphabet size above four for any K, L.

We derive new bounds on the total squared correlation (TSC) of quaternary (quadriphase) signature/sequence sets for all lengths L and set sizes K. Then, for all K, L, we design minimum-TSC optimal sets that meet the new bounds with equality. Direct numerical comparison with the TSC value of the recently obtained optimal binary sets shows under what K, L realizations gains are materialized by moving from the binary to the quaternary code-division multiplexing alphabet. On the other hand, comparison with the Welch TSC value for real/complex-field sets shows that, arguably, not much is to be gained by raising the alphabet size above four for any K,L. The sum-capacity (as well as the maximum squared correlation and total asymptotic efficiency) of minimum TSC quaternary sets is also evaluated in closed-form and contrasted against the sum capacity of minimum-TSC optimal binary and real/complex sets.

We study in this paper the sum capacity achievability of orthogonal transmissions in vector Gaussian multiple access channels (MAC). Specifically, we derive sufficient and necessary conditions, in terms of channel matrices and transmitter power constraints, for orthogonal transmissions to achieve the sum capacity of a vector Gaussian MAC. The obtained conditions provide a unified framework that helps explain many intuitive and known results as well as explore cases that have not been addressed. In the cases when these conditions are violated, our results enable us to quantify the suboptimality of orthogonal transmission when the sum capacity can only be achieved by overlay transmission.

We study in this paper the sum capacity achievability of orthogonal transmissions in vector Gaussian multiple access channels (MAC). Specifically, we derive the sufficient and necessary conditions, in terms of channel matrices and transmitter power constraints, for orthogonal transmission to achieve the sum capacity of a vector Gaussian MAC. The obtained conditions provide a unified framework in explaining many of the results that are intuitively true. They also enable us to explore cases that have not been addressed to determine the (sub)optimality of orthogonal transmissions compared with the overlay transmission.

We consider a multilayer perceptron neural network (NN) receiver architecture for the recovery of the information bits of a direct-sequence code-division-multiple-access (DS-CDMA) user. We develop a fast converging adaptive training algorithm that minimizes the bit-error rate (BER) at the output of the receiver. The adaptive algorithm has three key features: i) it incorporates the BER, i.e., the ultimate performance evaluation measure, directly into the learning process, ii) it utilizes constraints that are derived from the properties of the optimum single-user decision boundary for additive white Gaussian noise (AWGN) multiple-access channels, and iii) it embeds importance sampling (IS) principles directly into the receiver optimization process. Simulation studies illustrate the BER performance of the proposed scheme.

In this paper we prove formally that the optimum (nonlinear) DS-CDMA single-user decision boundary exhibits the following properties: (i) it is symmetric with respect to the origin and (ii) as it is traversed away from the origin, it converges to a hyperplane parallel to the MF decision boundary. Then, we translate properties (i) and (ii) to a set of constraints that can be used by any optimization algorithm for the selection (training) of the parameters of a general multi-layer-perceptron neural-network receiver. Using these constraints, the number of parameters to be optimized is reduced by nearly 50% for large-size networks, which effectively doubles the speed of any training procedure. Furthermore, we utilize properties (i) and (ii) to develop a new initialization scheme that provides additional improvements on the convergence rate and can be used by any recursive optimization algorithm. As a representative case study we consider the back-propagation (BP) algorithm and develop a constrained version of it that incorporates both the proposed constraints and the proposed initialization. The convergence rate enhancement achieved fay constrained-BP is illustrated by simulations.