We introduce a lightweight, multi-stage framework called Correlation-Enhanced CSI Sampling (CESS) to address this challenge. CESS combines full-slot channel interpolation, inter-slot CSI processing to close the temporal measurement gap, and intra-slot subcarrier processing to mitigate frequency-domain correlation issues. Through simulations, we compare CESS to a conventional PLKG baseline, showing that CESS substantially lowers the Key Disagreement Rate (KDR) and boosts the effective Key Generation Rate (KGR), with a minimum KGR of 3.5 kbps. These results underscore CESS’s potential for secure and reliable PLKG in real-world 5G TDD systems.

This paper studies time difference of arrival (TDoA)based algorithms for drone controller localization and analyzes TDoA estimation in multipath channels. Building on TDoA estimation, we propose two algorithms to enhance localization accuracy in multipath environments: the Maximum Likelihood (ML) algorithm, and the Least Squares Bancroft with GaussNewton (LS-BF-GN) algorithm. We evaluate these proposed algorithms in two typical outdoor channels: Wireless Local Area Network (WLAN) Channel F and the two-ray ground reflection (TRGR) channel. Our simulation results demonstrate that the ML and LS-BF-GN algorithms significantly outperform the LSBF algorithm in multipath channels. To further enhance localization accuracy, we propose averaging multiple tentative location estimations. Additionally, we evaluate the impact of time synchronization errors among sensors on localization performance through simulation.

Enhanced Mobile Broadband (eMBB) and Ultra Reliable Low Latency Communication (URLLC) are two important wireless communication traffics to build a digital twin of physical reality. Therein, eMBB and URLLC traffics are to transmit high-quality sensed data and critical commands, respectively. To support these two important traffics, we develop an intelligent resource allocation mechanism. First, we model the time-frequency resource allocation as an optimization problem aiming to maximize the throughput for the eMBB traffics according to their urgency subject to the constraint on the successful transmission for the URLLC traffics. In this way, the amount of resources allocated to each traffic can be appropriately determined without causing waste in resource usage. Secondly, we propose a feasible low-complexity solution for the optimization problem by relaxing it and then applying linear programming. Thirdly, to address the possible failure of the algorithm due to the relaxation, we propose a post-processing by puncturing the resource initially allocated to eMBB traffics and thereafter reallocating this resource to URLLC traffics. By such, the characteristic of the eMBB traffic, i.e., high throughput, and that of the URLLC traffic, i.e., low latency and ultra reliability, can be achieved. We perform system-level simulations on Matlab 5 G simulation platform to evaluate the performance of the proposed mechanism under different scenarios. Simulations show that the proposed mechanism achieves better performance compared to existing schemes regarding the total eMBB throughput and URLLC failure probability on all the scenarios.

We propose a scheme which generates shared secret keys from the amplitude of frequency domain channel state information (CSI) of individual orthogonal frequency-division multiplexing (OFDM) subcarrier. A unique quantization scheme is proposed, with the use of two parameters,$a$and β, to set the quantization threshold. Based on multivariate Rayleigh distribution, we derive the analytical expressions for key bit disagreement rate (BDR) and key length. The theoretical analysis of BDR and key length aligns with the simulation results. We also present a mathematical model to optimize the parameters of our proposed quantization scheme and a guideline to set$a$and β. Our proposed method for iterative tuning of β can balance the number of zeros and ones in the generated key sequences so that the randomness requirement is met.

In this work we investigate the design of energy-optimal policies for ultra-reliable low-latency communications (URLLC) over a non-stationary wireless channel, using a contextual reinforcement learning (RL) framework. We consider a point-to-point communication system over a piece-wise stationary wireless channel where the Doppler frequency of the channel switches between two distinct values, depending on the underlying state of the channel. To benchmark the performance, first we consider an oracle agent which has a perfect but causal information about the switching instants, and consists of two deep RL (DRL) agents each of which is tasked with optimal decision making in a unique partially stationary environment. Comparing the performance of the oracle agent with the conventional DRL reveals that the performance gain obtained using oracle agent depends on the dynamics of the non-stationary channel. In particular, for a non-stationary channel with faster switching rate the oracle agent results in approximately 15 − 20% less energy consumption. In contrast, for a channel with slower switching rate the performance of the oracle agent is similar to the conventional DRL agent. Next, for a more realistic scenario when the information about the switching instants for the Doppler frequency of the underlying channel is not available, we model the non-stationary channel as a regime switching process modulated by a Markov process, and adapt the oracle agent by aiding a state tracking algorithm proposed for the regime switching process. Our simulation results show that the proposed algorithm yields a better performance compared to the conventional DRL agent.

Radio interference or jamming can cause isolated area in advanced metering infrastructure (AMI) based on Wire-less smart utility network (Wi-SUN), in which conventional recovery techniques cannot cope with. In this paper, we deploy narrowband internet-of-things (NB-IoT) interface to some smart meters to act as emergency gateway, called ResiLite. An optimal ResiLite placement algorithm for enhancing network resilience is proposed. We define an implicit resilience metric based on path diversity and cluster closeness. The defined metric is used to form an integer linear programming (ILP) problem, then we solve the ILP to obtain optimal ResiLite placement. Simulation results show that the optimal ResiLite placement improves Wi-SUN AMI network resilience (defined as the number of surviving nodes with packet delivery ratio (PDR) above 99% under disturbance) by up to 167% compared to an AMI without ResiLite, and up to 29% compared to uniformly random ResiLite placement, respectively.

A practical power amplifier (PA) has nonlinear characteristics that distort the output signal and hence increase the transmission error. Digital pre-distortion (DPD) has been widely accepted to compensate for the PA nonlinearity. However, in direct-conversion transmitters (DCTs), DPD performance is affected by in-phase and quadrature (IQ) imbalance. In this paper, we utilize the Iterative Learning Control (ILC) algorithm to design a DPD scheme to compensate for PA nonlinearity under IQ imbalance. We first prove that ILC is applicable in such a scenario. This proof is validated using simulations which show that ILC is able to estimate the PA ideal input. The estimated ideal input is then exploited in training a neural network (NN)-based DPD model. We provide the complexity estimation of our proposed scheme using the number of real multiplications. Finally, we demonstrate the performance advantage of our proposed scheme in comparison with other existing polynomial based approaches through simulations and measurements.

Advanced metering infrastructure (AMI) based on Wireless Smart Utility Network (Wi-SUN) employs carrier sense multiple access (CSMA), hence suffers from poor network performance when the number of nodes increases but without proper design of data collection scheduling. In this paper, we introduce an interference-aware TDMA-like optimal data collection scheduling, in which a link-timeslot assignment problem is formed by considering interference constraint to achieve better spatial timeslot reuse. A constraint for achieving consecutive flow for multi-hop transmission is also introduced to avoid modifications on the original CSMA protocol, and hence the proposed scheduling is standard compliant. Our results show that the proposed scheduling reduces the total data collection time of Wi-SUN AMI network with 100(200) smart meters by 59%(48%), 33%(32%), and 19%(13%), respectively, when compared with conventional Wi-SUN AMI, conventional TDMA scheduling and heuristic scheduling based on 2-rank distance interference model.

Accurate wireless interference identification (WII) is vital for wireless industrial internet of things (IIoT) network to coexist with other technologies in the crowded 2.4 GHz unlicensed band. Deep learning (DL) based methods have emerged as a promising candidate for such type of task. However, to achieve good accuracy, DL methods require large amount of labeled training data, which comes from tedious annotation work by domain expert. In contrast, unlabeled data is easier to obtain. In this paper we present a semi-supervised DL based WII algorithm which combines temporal ensembling technique with CNN network to exploit unlabeled data to improve the performance. The proposed algorithm is able to differentiate interference from multiple wireless standards accurately with reduced number of labels, such as IEEE 802.11, IEEE 802.15.4 and IEEE 802.15.1. Specifically, the proposed algorithm achieves 90% accuracy with less than 2% of labeled data with medium to high signal SNR. Extensive simulation results show that the proposed algorithm achieves a better classification accuracy than benchmark algorithms under various SNR conditions and with different number of labeled data.

We evaluate the performance of machine learning method in identifying the channel type in 802.11ac systems. It is shown that the reference symbols contained in the packet preamble can be used as a good feature for classification. In addition, the time-domain received preamble also serves as a good classification feature with comparable performance, allowing early identification of channel type since the information can be tapped closer to the receiving antenna. We validate our approach in both software and hardware simulation, and classification accuracy of more than 94% is shown to be attainable at moderate Signal to Noise Ratio (SNR). Finally, we also evaluate the application of our proposed algorithm into link adaptation in 802.11ac systems, and show that up to 1.6dB gain can be achieved compared to the case without channel type identification.

In IIoT application, various IoT devices with different protocols are connected. One of the major challenges in IIoT application is to detect interference and achieve coexistence across multiple technologies. In this paper, we propose a novel and efficient method to detect the presence of Microwave Oven (MWO) interference. In order to achieve coexistence between IEEE 802.11 WIFI and IEEE 802.15.4 based sensor network, we propose an enhanced spectral matching method to detect 802.15.4 packet and WIFI packet. Our proposed method can also detect the hopping frequency of 802.15.4 packet. Simulation results show that our proposed method can achieve high detection probability in various propagation channels.

In this paper, we aim to design and optimize the performance of IEEE 802.11ad-based dense network in cabin environment. For a typical airplane model, we propose a suitable network deployment and build up the directional antenna model and path loss model. We also develop an electromagnetic algorithm to predict the in-cabin 60 GHz radio propagation characteristics and the channel model. Considering the blockage effects, we further investigate the link level and system level performance of IEEE 802.11ad networks for different modulations and different cabin regions. The simulation results validate that the IEEE 802.11ad network is applicable to the wireless in- flight entertainment and communication (WIFEC) services. Moreover, the adopted rate adaption scheme dramatically improves the robustness of IEEE 802.11ad networks, which achieves better performance than fixed modulation scheme.

In this paper we propose a Hybrid Group Paging scheme to overcome radio access networks (RANs) overload problem due to simultaneous access from massive number of machine-type communications (MTC) devices. Compared to the existing methods, the proposed scheme supports dynamic priority assignment among different sub-groups by combining features from both pull-based and push-based approaches. The performance of the proposed scheme is analyzed using recursive contending user estimation (RCE) method, and shown to agree well with the simulation results.

In this paper, we present a user equipment (UE) based distributed traffic steering mechanism between long-term evolution (LTE) and Wi-Fi networks. An agent residing in each UE evaluates the traffic condition of the network it is currently connecting to and makes the traffic steering decision. The evaluation is either periodic or event-driven such as access denial in the admission control due to network congestion. The learning mechanism enables each UE to use the locally available information at the UE and select the proper network under dynamic network conditions. The computer simulation results show that the proposed mechanism achieves low outage probability and small number of network switching with even less information than or almost the same as the existing method. We have also implemented the proposed traffic steering mechanism as an APP on android platform and verified that the proposed mechanism works effectively in real-time testing.

The provision of Quality of Experience (QoE) for wireless video streaming has become a necessity. From this perspective, we design a QoE-aware scheduling (QoEAS) scheme for video streaming over IEEE 802.11n/ac networks. We consider an application scenario where videos are streamed in an area with high density of users. To improve the streamed video quality, we enhance the conventional gradient based scheduling scheme (U'R) by incorporating the packet delay and channel transmission rate with video packet importance index in scheduling. This index is derived based on the gradient of the QoE function for video, which accurately reflects the usefulness of a packet in enhancing the video quality. To be compatible to practical networks such as IEEE 802.11n/ac, the packet importance index is embedded in the Type-of-Service (ToS) field of IP header and is employed at Media Access Control (MAC) layer to perform scheduling. The proposed QoEAS significantly enhances the streamed video quality and achieves good fairness among users. It surpasses the conventional scheme by as much as 5 dB in Peak Signal-to-Noise Ratio (PSNR) at the lower 10 percentile of the users. To demonstrate the superiority and practicability of QoEAS, we also set up a lab test bed on a real 802.11ac network. Compared to conventional schemes, the proposed QoEAS shows notable QoE enhancement in lab test.

In this paper, we will first motivate the heterogeneous network as an evolutionary path to the fifth generation (5G) communications. We then present an agile software defined heterogeneous network architecture which virtualizes the various radio access networks such as cellular basestations and Wi-Fi access points and the various carrier frequencies from both licensed and unlicensed bands. The software defined heterogeneous network architecture can therefore support much more efficient resource utilization and meet the quality of service requirement. We will also share our work in context-aware Wi-Fi-cellular network traffic steering and mobility management as two use cases in the software defined heterogeneous network.

We propose a opportunistic multicast scheduling scheme to exploit content reuse when there is asynchronicity in user requests. A unicast transmission setup is used for content delivery, while multicast transmission is employed opportunistically to reduce wireless resource usage. We then develop a multicast scheduling scheme for the downlink multiple-input multiple-output orthogonal-frequency division multiplexing system in IEEE 802.11 wireless local area network (WLAN). At each time slot, the scheduler serves the users by either unicast or multicast transmission. Out-sequence data received by a user is stored in user's cache for future use. Multicast precoding and user selection for multicast grouping are also considered and compliance with the IEEE 802.11 WLAN transmission protocol. The scheduling scheme is based on the Lyapunov optimization technique, which aims to maximize system rate. The resulting scheme has low complexity and requires no prior statistical information on the channels and queues. Furthermore, in the absence of channel error, the proposed scheme restricts the worst case of frame dropping deadline, which is useful for delivering real-time traffic. Simulation results show that our proposed algorithm outperforms existing techniques by 17 % to 35 % in term of user capacity.

We study the optimality of separate network-channel coding (SJNC) for three independent sources on the orthogonal access scheme. By SJNC, we mean that the network coded messages are formed from the binary XOR of the sources' messages, followed by channel encoding which is independent of the network coding. Decoding is performed jointly across the network and channel codes. Using standard random coding and joint typical set decoding, we obtain an achievable rate region for SJNC. By allowing time-sharing of codes, the rate region coincides with the capacity region of joint network-channel coding, where network and channel coding are performed jointly. This proves, surprising, that there is no loss of optimality for SJNC.

In many communication scenarios, such as in cellular systems, the energy cost is substantial and should be conserved, yet there is a growing need to support many real-time applications that require timely data delivery. To model such a scenario, in this paper we consider the problem of minimizing the expected sum energy of delivering a message of a given size from a source to a destination subject to a deadline constraint. A relay is present and can assist after it has decoded the message. Causal channel state information (CSI), in the form of present and past SNRs of all links, is available for determining the optimal power allocation for the source and relay. We obtain the optimal power allocation policy by dynamic programming and explore its structure. We also obtain conditions for which the minimum expected sum energy is bounded given a general channel distribution. In particular, we show that for Rayleigh and Rician fading channels, relaying is necessary for the minimum expected sum energy to be bounded. This illustrates the fundamental advantage of relaying from the perspective of energy efficient communications when only causal CSI is available. Numerical results are obtained which show the reduction in the expected sum energy under different communication scenarios.

We propose a time slot based transmission strategy, referred to as adaptive coordinated napping (CoNap), for energy saving in cellular networks under time-varying traffic demand. In adaptive CoNap network, multiple neighboring base stations (BSs) form a cluster and each BS operates in either a transmit mode (TM) or a nap mode (NM) in each time slot. The dynamic assignment of TM and NM to each BS is implicitly coordinated among multiple BSs. This implicit coordination is realized by a binary general flickering pattern matrix (FPM) through adaptively selected mapping matrix (MM). To track the time-varying traffic demand, we develop an adaptive algorithm to dynamically select the appropriate MM from a predefined MM set by taking into account the network quality of service (QoS) requirement. Our numerical results based on a realistic energy consumption model in a cellular network show that as high as 40% saving can be achieved without compromising the network QoS.

In this paper, we employ a relay to assist in transmission from a source to a destination over a fixed number of time slots. We seek to minimize the expected sum transmission energy used by the source and relay, by choosing the transmission power of each node dynamically slot by slot. We assume that both receivers, namely the relay and the destination, can accumulate mutual information across previous and present slots by using earlier received packets for joint decoding. Assuming availability of causal channel state information (CSI) at all transmitters, we obtain the structure of the optimal solutions by dynamic programming. A key insight is that relaying is necessary for the expected sum energy to be bounded in Rayleigh fading channels. Numerical results show that the reduction in the expected sum energy is significant.

The key of the multimedia compression technique is trying to remove the correlation between the neighboring frames. Due to the computational complexity limitation, considerable correlation is still observed in the multimedia-coded sequence. This residual redundancy can be modeled as a first-order Markov model with the transition probability of the multimedia-coded parameter between adjacent frames. Therefore, in the multimedia transmission over wireless network, there have been researches on the iterative joint source-channel decoding (ISCD), using the transition probability as source side information (SSI) to reduce errors due to the noisy channel. To further correct the errors after the ISCD, we propose a low-complexity yet efficient bit-flipping (BF) algorithm where restrictions are imposed to prevent incurring high undetected error frames. For the ISCD-BF, the source transition probability is crucial to its performance. However, in practice, it is unknown to the receiver. Thus, we propose an SSI estimation to update the transition probability from the output of the ISCD-BF. The performances of the proposed algorithms are examined by applying to a speech transmission over a wireless network. The simulation results show that our proposed algorithms achieve notable gain over the conventional decoder.

We consider a time-slotted source-relay-destination network where data is to be delivered by a deadline. Our goal is to minimize the sum transmission energy by power allocation for each node and time slot, with knowledge of full channel state information (CSI) in the form of the SNRs of all slots before the deadline. We assume that both receivers, namely the relay and the destination, can accumulate mutual information across previous and present slots by using earlier received packets for joint decoding. We obtain the structure of the optimal solution in a semi-analytical closed-form. The structural result shows that the source and relay jointly perform a generalized form of water-filling over slots. Finally, we use the structural result to develop a heuristic scheme that uses only causal CSI, in the form of the SNRs of only past and present slots.

We consider an energy efficient resource allocation method for orthogonal frequency division multiple access (OFDMA) systems consisting of M users and N subchannels. Transmit power assignment follows the optimal strategy, i.e., a water-filling strategy, for the given subchannels. For the subchannel allocation, we introduce a recently proposed algorithm called a sequential linear assignment algorithm (SLAA). The SLAA determines how many subchannels are allocated to each user and which subchannels are allocated to each user by solving the outer and inner problems, respectively, in an iterative manner; as a result, it requires O(MN2(N-M)2) complexity. To reduce the computational complexity of SLAA, we propose a simplified sequential linear assignment algorithm (SSLAA). Based on the observation that a user requiring the higher power would take the more subchannels, SSLAA uses each user's power consumption as a metric to determine who will take an additional subchannel in each iteration. Consequently, contrast to SLAA using network power consumption as a metric, the SSLAA incurs significant reduction of computational complexity, by O(N2(N-M)2). Computer simulations in compared with other existing heuristic algorithms verify that the proposed SSLAA results in high energy efficiency and low computational complexity.

In this paper, we consider the half-duplex cooperative multiple-access channel (CMAC) with frequency-selective block-fading. Each link employs an orthogonal frequency division multiplexing (OFDM) system, where modulated symbols are drawn from a finite constellation set. We first obtain the diversity order of the CMAC, as a function of the time sharing variables of the users and the rates of the codes. To achieve this rate-diversity tradeoff, we use the principle of network coding where messages of the two sources are jointly encoded. Both separate and joint network-channel coding approaches are considered. Specifically, we design multiple turbo codes that minimize the outage probabilities of these approaches. We also give a code structure for the multiple turbo codes to achieve full diversity of the system. The codes are optimized using the extrinsic information transfer (EXIT) chart analysis with iterative decoding tailored for OFDM modulated CMAC. Numerical examples show that with our proposed design technique, the achieved frame error rate is within 0.5dB from the information outage. Without network-channel coding, the outage probability of distributed coding cannot achieve the diversity order given in the rate-diversity tradeoff.

The encoding and decoding of network coding at network layer and channel coding at the physical layer can be performed either jointly or separately. The largest achievable rate region is obtained when both are done jointly. Our interest in this paper lies in investigating the optimality of separate encoding and joint decoding of network-channel (NC) code (SJNC) assuming an orthogonal access scheme. Our key result is that SJNC is optimal as it achieves the capacity region of JJNC. However, SJNC is suboptimal if time sharing of codes is not allowed, or if separate decoding is used. We then apply these results to the cooperative multiple-access channel (CMAC) where time-sharing of codes is practically infeasible. Optimality of SJNC is compared to JJNC in terms of outage region. The conditions for full diversity order of orthogonal frequency division multiplexing (OFDM) modulated CMAC with JJNC and JSNC are then derived. Using numerical examples, multiple turbo codes are designed to show that the achieved frame error rate (FER) is within 0.6dB from the information outage.

In this paper, we employ a relay for scheduling with deadline, i.e., a relay assists to deliver a given number of bits from a source to a destination over a fixed number of slots. Prior to every transmission, the channel of the present slot is made known to all nodes, but not of future slots. We seek to minimize the sum energy used, by choosing the transmission power and transmission duration of each node dynamically slot by slot. Assuming a decode-and-forward scheme for every slot, we obtain the optimal policy via dynamic programming. We also propose and analyze an asymptotically optimal policy that can be solved as a series of convex optimization problems. Numerical results show that the proposed policy gives a tight upper bound on the minimum sum energy, and that a significant fraction of energy can be saved with relaying.

In this paper, we consider the half-duplex relay channel. We seek to design multiple turbo codes to minimize the information outage in the block fading channel. An analysis on the diversity order of the relay channel, which depends on the time sharing variable and the rate of the code, is given for practical modulations. We also give a code structure for the multiple turbo codes to achieve the full diversity when it is achievable. The codes are optimized using the extrinsic information transfer (EXIT) chart analysis, based on the convergence thresholds of the iterative decoding tailored for relay channel. Numerical examples shows that with our design technique, the achieved frame error rate is within 0.7dB of the information outage. To reduce complexity, a suboptimum code search approach is proposed, resulting codes which perform 1dB away from the information outage.

A low complexity M-algorithm based multiple-input multiple-output (MIMO) tree search algorithm with near maximum likelihood (ML) performance is proposed in this paper. Numerical examples show that our tree search algorithm is able to provide a significant performance gain over the MMSE detection. Based on this algorithm, a fully pipelined architecture is presented for the MIMO orthogonal frequency division multiplexing (OFDM) systems. The throughput for a 4x4 MIMO-OFDM IEEE 802.11n system with 64-QAM is 312 Mbps.

In this paper a simplified parallel belief propagation (BP) algorithm is suggested as a suboptimal soft-input soft-output multiuser detector in an iterative multiuser decoding scheme. The iterative decoding scheme itself is based on applying an outer BP algorithm for message passing between the BP multiuser detector and the single-user decoders. The performance of the iterative decoding process is investigated based on parallel scheduling of the outer BP algorithm. In addition, we analyze the large system BER performance of the corresponding multiuser decoding algorithms. By making Gaussian assumptions on the output of the single-user decoders and the BP multiuser detector, we derive expressions which describe approximately the BER performance of the algorithms. Finally, numerical examples are presented, demonstrating the accuracy of this approximation

In this paper, a class of nonlinear minimum mean-squared error multiuser detectors is derived based on a multivariate Gaussian approximation of the multiple-access interference for large systems. This approach leads to expressions identical to those describing the probabilistic data association (PDA) detector, thus providing an alternative analytical justification for this structure. A simplification to the PDA detector based on approximating the covariance matrix of the multivariate Gaussian distribution is suggested, resulting in a soft interference-cancellation scheme. Corresponding multiuser soft-input, soft-output detectors delivering extrinsic log-likelihood ratios are derived for application in iterative multiuser decoders. Finally, a large-system performance analysis is conducted for the simplified PDA, showing that the bit-error rate (BER) performance of this detector can be accurately predicted and related to the replica method analysis for the optimal detector. Methods from statistical neurodynamics are shown to provide a closely related alternative large-system prediction. Numerical results demonstrate that for large systems, the BER is accurately predicted by the analysis and found to be close to optimal performance.

It has been shown that the stable fixed points of belief propagation (BP) algorithms correspond to extrema of the Bethe free energy. In this paper, we describe the dual problem for the minimization of the Bethe free energy and solve it using simple nonlinear block Gauss-Seidel and Jacobi algorithms. The use of the nonlinear block Gauss-Seidel algorithm corresponds to serial scheduling for the BP algorithm. In addition, it is shown that applying the nonlinear block Jacobi algorithm on the dual of the Bethe free energy corresponds to the parallel BP algorithm.

We consider a multiuser detector based on modelling the multiple-access interference (MAI) as a vector of Gaussian random variables. This approach leads to the probabilistic data association (PDA) multiuser detector, which outputs good approximations to marginal posterior-mode optimal decisions. A simplification to the PDA detector is suggested, leading to a soft interference cancellation scheme. In the large system limit, the bit error rate performance of this detector can be accurately predicted, and shown to be identical to the optimal detector.

In this paper, we compare several optimization methods for solving the optimal multiuser detection problem exactly or approximately. The purpose of using these algorithms is to provide complexity constraint alternatives to solving this nondeterministic polynomial-time (NP)-hard problem. An approximate solution is found either by relaxation or by heuristic search methods, while the branch and bound algorithm is used to provide an exact solution. Simulations show that these approaches can have bit-error rate (BER) performance which is indistinguishable from the maximum likelihood performance. A tabu search method is shown to be an effective (in terms of BER performance) and efficient (in terms of computational complexity) heuristic when compared to other heuristics like local search and iterative local search algorithms. When the number of users increases, the tabu search method is more effective and efficient than the semidefinite relaxation approach.

A detection strategy based on a semidefinite relaxation of the CDMA maximum-likelihood (ML) problem is investigated. Cutting planes are introduced to strengthen the approximation. The semidefinite program arising from the relaxation can be solved efficiently using interior point methods. These interior point methods have polynomial computational complexity in the number of users. The simulated bit error rate performance demonstrates that this approach provides a good approximation to the ML performance.

The detection strategy usually denoted optimal multiuser detection is equivalent to the solution of a (0, 1)-constrained maximum-likelihood (ML) problem, a problem which is known to be NP-hard. In contrast, the unconstrained ML problem can be solved quite easily and is known as the decorrelating detector. In this paper, we consider the constrained ML problem where the solution vector is restricted to lie within a closed convex set (CCS). Such a design criterion leads to detector structures which are ML under the constraint assumption. A close relationship between a sphere-constrained ML detector and the well-known minimum mean square error detector is found and verified. An iterative algorithm for solving a CCS constraint problem is derived based on results in linear variational inequality theory. Special cases of this algorithm, subject to a box-constraint, are found to correspond to known, nonlinear successive and parallel interference cancellation structures, using a clipped soft decision for making tentative decisions, while a weighted linear parallel interference canceler with signal-dependent weights arises from the sphere constraint. Convergence issues are investigated and an efficient implementation is suggested. The bit-error rate performance is studied via computer simulations and the expected performance improvements over unconstrained ML are verified.

It has previously been shown that well known iterations for solving a set of linear equations correspond to linear interference cancellation structures. Here, we suggest applying a block-wise iteration that consists of an outer and an inner iteration. The outer iteration used is the Gauss-Seidel (GS) method, while for the inner iteration, we study direct matrix inversion, the Jacobi over-relaxation iteration, and the conjugate gradient iteration. When a true inner iteration is used, this approach allows for a timely derivation of the acceleration parameters required by advanced iterations. The block iteration is based on a symbol-level implementation which leads to the same detection delay profile for both parallel and serial structures at the expense of differences in the amount of serial processing required. This is discussed in some detail and quantified for comparison. The performance of the detectors is studied via computer simulations where it is found that the block approach can provide significantly faster convergence, leading to improved detection delay over the simpler GS iteration. The improvements are obtained at the expense of an increase in the required serial processing speed.