Semantic communication has been regarded as a promising technology to serve upcoming intelligent applications. However, few studies have addressed the problem of resource allocation in semantic communication networks. Most resource allocation mechanisms act fairly to all original data, ignoring the meaning behind the transmitted bits. In this paper, a dynamic resource allocation scheme for the task-oriented semantic communication network (TOSCN) based on deep reinforcement learning (DRL) is proposed, which allows data with richer semantic information to preferentially occupy limited communication resources. This paper aims to design a deep deterministic policy gradient (DDPG) agent at the micro base station to maximize the long-term transmission efficiency of tasks. Firstly, the relationship between semantic information and task performance is investigated. Subsequently, a novel wireless resource allocation model for TOSCN is proposed by taking the image classification task as an example. Then, a joint optimization problem of the semantic compression ratio, transmit power, and bandwidth of each user is formulated. The agent is trained in an interactive learning environment to obtain a decent trade-off between the amount of data delivered to the receiver and the accuracy of intelligent tasks. Simulation results demonstrate that the proposed scheme achieves significant advantages in relieving communication pressure and improving task performance in resource-constrained wireless networks.

This paper studies the unmanned aerial vehicle (UAV) based multiple intelligent reflecting surface (IRS) network, where the hovering UAV acts as a base station, and the IRS enhances signal transmission to across obstacle between users and UAV. To achieve the maximum sum rate of proposed communication scenario, a non-convex problem considering IRS association results, hovering altitude of UAV, and the phase shift design of multi-IRS is formulated. From the IRS association problem, we can find that the IRS association results are coupled to the decoding order of non-orthogonal multiple access (NOMA). To tackle this, a mathematical interference expansion scheme is developed to decouple it and transform it to convex by binary relaxation method. The non-convexity of hovering altitude optimization problem is solved by logarithm operation, approximation, and auxiliary matrices. For the phase shift optimization problem of multi-IRS, we propose a gradient approximation based initial scheme and develop a univariate optimization based approach on the basis to achieve the users sum rate improvement in multi-IRS. In the end, we compare the proposed scheme with baseline scheme to present the superiority of this work under various network settings. The internal reasons for the variation of simulation results are also analyzed.

Driven by the practicality of unmanned aerial vehicle (UAV), we consider a dual-UAV based non-orthogonal multiple access (NOMA) scenario, which consists of one communication UAV for services and one jamming UAV against eavesdropping. The goal is to maximize the secrecy energy efficiency through the successive convex approximation based communication resource, UAV trajectory, and artificial noise optimization. Considering the probabilistic constraint of outage probability from imperfect channel state information, we transform it to a non-probabilistic problem by Markov inequality and Marcum$Q$-function, then the problem is decomposed into three subproblems. We apply matching-swapping method to assign subchannel in non-orthogonal multiple access (NOMA) UAV networks before the joint process, then convert the power optimization problem to a standard convex optimization form by upper bound of the concave function. The communication UAV trajectory is studied under the constraints of flying energy consumption, maximum speed, and flying altitude. To track this NP-hard problem, Taylor expansion and various slack variables sets are introduced to transform the non-convex problem to convex one. For the artificial noise optimization problem, we use the lower bound to replace the convex term turning it into an easy-to-solve convex optimization problem. In the end, simulations results reveal that: 1) The reasonable jamming scheme can improve the secrecy energy efficiency of the NOMA UAV networks, even if it can cause interference for legitimate users; 2) UAV will fly to a place where the performance gain from users is high when flying energy consumption permits.

Replacing base stations with unmanned aerial vehicles (UAVs) to serve the communication of ground users has attracted a lot of attention recently. In this paper, we study the joint resource allocation and UAV trajectory optimization for maximizing the total energy efficiency in UAV-based non-orthogonal multiple access (NOMA) downlink wireless networks with the quality of service (QoS) requirements. To handle the user scheduling problem, a heuristic algorithm based on matching and swapping theory is proposed first to allocate users that access UAV in each subperiod, then the transmit power allocation problem which considers the maximum transmit power and minimum user date rate is transformed to a convex optimization problem using logarithmic approximation. Meanwhile, the successive convex optimization is used in UAV trajectory optimization problem and a joint optimization algorithm is presented with the algorithm’s convergence and computational complexity. Finally, numerical results are provided to support the rationality of the proposed algorithm.

In this paper, we mainly studied the joint resource allocation and UAV trajectory optimization for maximizing the total energy efficiency in UAV-based non-orthogonal multiple access (NOMA) downlink wireless networks with the quality of service (QoS) requirement. To track the joint optimization problem, a heuristic algorithm based on matching theory in cellular networks is proposed firstly to allocate users which connect the UAV in each subperiod, then the transmit power allocation problem which considers the maximum transmit power and minimum user date rate is transformed to a convex optimization problem by logarithmic approximation, and solved to get an optimal solution. Meanwhile, the successive convex optimization is used in UAV trajectory optimization problem for its near-optimal solution. Finally, numerical results are provided to support the rationality of the proposed algorithm.

This paper intends to study the energy efficiency in caching based UAV networks, where fog radio access network (FRAN) and non-orthogonal multiple access (NOMA) are considered meanwhile. Taking full account of the impact of caching, subchannel assignment, and power allocation in UAV enabled wireless networks, we formulate the problem of maximizing energy efficiency. In order to better solve the proposed non-convex problem, we propose a subchannel assignment algorithm and a power allocation algorithm applying alternating direction method of multipliers (ADMM). The final simulation section verifies the fast convergence of the algorithm and compares the advantages with existing algorithms.

In this paper, an improved full rank condition (named as λ-FRC) is proposed for the construction of nonbinary LDPC codes. Compared with the conventional FRC, the λ-FRC introduces a metric λ to evaluate the reliability of the messages passing through the circle. Codes with better performance can be obtained by selecting elements of the parity matrix according to the metric. Based on the λ-FRC, a greedy search algorithm is proposed to design the codes efficiently. In order to further improve the performance of the codes constructed, an iterative search algorithm is applied in the construction. Simulation results show that the performance of codes constructed by the λ-FRC methods outperforms those by the FRC methods, and the iterative search algorithm improves the performance of the codes as the iteration number increases1.

Vector Orthogonal Frequency Division Multiplexing (V-OFDM) for single transmit antenna systems is a generalization of OFDM where Single-Carrier Frequency Domain Equalization (SC-FDE) and OFDM are two special cases. It has been shown to be able to collect multipath diversity and thus generally outperforms OFDM. The performance of OFDM under phase noise has been extensively studied in literature. These effects can be summarized as introducing a Common Phase Error (CPE) and Inter Carrier Interference (ICI) resulting in poor performance. The performance of V-OFDM under phase noise still remains uninvestigated and in this paper it is shown that phase noise in V-OFDM systems leads to a Common Vector Block Phase Error (CVBPE) as well as an Inter Vector Block Carrier Interference (IVBCI) effect. Both of these effects are detrimental to the performance of a V-OFDM system. An exact expression for the Signal to Interference Noise Ratio (SINR) for a V-OFDM system is derived and found to agree closely with simulation results.

In-phase and Quadrature Imbalance (IQI) exists in both analog IQ modulator and demodulator. When carrier frequency is ultra-high and/or bandwidth is ultra-wide, the impact of IQI on system performance is not negligible. While in most of the literatures, the combined effect of channel, TX, and RX IQI is discussed, in this paper, a method that can separate and simultaneously estimate the three is proposed. The complexity of the proposed method is low, and with estimated TX and RX IQI, simple symbol-by-symbol detection can be used to detect data symbols with negligible performance loss. Compared with the calibration methods, where auxiliary circuits are needed to separate the TX and RX IQI belonging to different communication links inside a transceiver, by exploiting the fact that the wireless channel changes much faster than the IQI, the method proposed here can separate not only the TX and RX IQI but also the channel belonging to the same communication link without any auxiliary circuits. Simulations are carried out to verify the performance of the proposed method.

Using large number of antennas at base station (BS) to formulate a multi-user multiple-input multiple-output (MU-MIMO) system, i.e., the massive MIMO, has attracted more and more attention recently, due to the high spectrum and energy efficiency it has. When there is correlation and mutual coupling, Rusek et al. showed that the channel may lose asymptotic orthogonality as the number of antennas increases. In this paper, we investigate the dual-polarized (DP) antenna array with exponential correlation and coupling model and prove that under the assumption that the correlation and coupling decreases exponentially with the distance between the antennas, the channels are still asymptotically orthogonal and we still have the same massive MIMO effect with the single-polarized (SP) antenna array. Through simulations, the asymptotic expressions are verified and the benefits brought by DP antenna array in massive MIMO system are quantitatively shown for both uniform linear array (ULA) and uniform planar array (UPA) configurations.

In order to achieve reliable communication, transmit diversity is commonly used in multiple antenna systems. Currently, to obtain transmit diversity, space-time code is usually used. In this paper, a new multiplicative transformation scheme is proposed for multiple antenna systems with non-binary low-density parity-check (LDPC) codes. It is shown that this scheme can provide transmit diversity. In the scheme, for the first antenna, the modulated LDPC codeword is directly applied, while for the second antenna, before modulation, a multiplicative transformation is used, where the multiplication is defined in the same Galois field as that used to construct the nonbinary LDPC code. The performance of the proposed system is analyzed, and it is shown that, compared with phase sweeping transmit diversity (PSTD) and Alamouti's space-time block code (STBC), the proposed system not only provides transmit diversity, but also achieves good gain. Simulation results further prove the increased performance in both block fading and slow fading channels.

In this letter, the mathematical equivalence between the vector OFDM proposed by Xia in 2001 and the quadrature OFDMA proposed by Zhang et. al. in 2009 is shown. Having the mathematical equivalence, all the discussed aspects of the one can be applied to the other in a straightforward manner.

In this paper, we study the transmission strategy in a two-way relay channel. In particular, we propose a new class of logM1(M2)-MA (multi-access) block relay protocols where an M1-ary modulation is used in the MA phase, and an M2-ary in the broadcast phase. Lower-order (e.g., BPSK) and higher-order (e.g., 16-QAM, 32-PSK) modulations are used respectively for the multi-access phase and the broadcast phase, in contrast to the conventional method of using the same modulation type (of either higher or lower order) throughout. To achieve this, the relay buffers the received symbols from consecutive multi-access phases and performs signal combining in conjunction with the decode-and-forward protocol. We derive the theoretical performance bounds for the Rayleigh fading channels and closed forms for the AWGN channels. Numerical results confirm the advantages of the proposed scheme with an improved throughput in lower signal-to-noise ratio range.

In MIMO broadcast channels, the Block Diagonalization (BD) transmit precoding performs better than the Zero-Forcing (ZF) transmit precoding when the user equipment has multiple receive antennas and can support multiple data streams. The number of users simultaneously supported by BD is much less than the total number of users in the system, so selecting proper users and/or receive antennas is important to achieve high overall sum capacity. So far, the capacity-based and norm-based user and receive antenna selection algorithms are proposed, which are greedy based and have much lower complexity than the brute-force based algorithms. In this paper, the orthogonality-based user and receive antenna selection algorithms are employed. The idea of the algorithm is simple, however, we show by detailed complexity analysis and extensive simulations that the orthogonality-based user and receive antenna selection achieves much better complexity and performance tradeoff than the existing capacity-based and norm-based algorithms. So, the orthogonality-based algorithms are more attractive for practical MIMO broadcast systems.

In the paper, an improved link adaptation scheme is proposed for the WiMedia MB-OFDM UWB system, in which quality of service (QoS) support is provided. The proposed scheme consists of three functional blocks: link quality indicator (LQI) calculator, frame error rate (FER) estimator, and transmitter (TX) parameter selector. Instead of using the average receive SNR (ASNR) as LQI, a new LQI metric is defined based on a union bound analysis to capture the effects of both path loss and frequency selectivity of an instantaneous UWB channel. How to calculate LQI for each rate mode is investigated by analyzing the distribution of soft bit information. With the calculated LQI, the FER performance of each rate mode can be accurately estimated with a look-up table method, which is suitable for practical implementation. Using the estimated FERs, TX parameter selector can optimize TX rate mode to improve the communication throughput under QoS constraints. It is shown the proposed scheme can significantly improve the throughput while maintaining the required QoS compared to conventional ASNR based link adaptation schemes.

As a space- and cost-efficient antenna configuration, dual-polarized antennas are widely deployed in real field MIMO wireless communication systems. The Quasi-Orthogonal Space-Time Block Code (QOSTBC), due to its advantages in the transmission rate and the decoding complexity, is an important transmit diversity scheme for more than 2 transmit antennas. In this paper, we investigate the performance of QOSTBC for dual-polarized MIMO channels. We show that when QOSTBC is used with dual-polarized antennas, we need to connect the power amplifiers (PAs) to the differently polarized antennas carefully, in order to avoid more than 1dB performance loss at high cross-polar discrimination (XPD). Also, we show that, if the QOSTBC with constellation rotations, which achieves full diversity in uni-polarized MIMO channels, is used, the performance loss due to the bad PA-to-antenna connections is mitigated. However, when the XPD is high, the QOSTBC without full diversity, if connected properly, can achieve better performance than the QOSTBC with full diversity when connected badly, which shows the importance of a proper connection at high XPD. So, to guarantee robust performance of QOSTBC at any XPD, besides designing codes to have full diversity and good diversity product, connecting the PAs and the antennas properly is also important.

This paper considers the time and frequency synchronization based on the interspersed pilots. The shortcoming of an existing algorithm is observed, and a new channel estimation assisted time and frequency algorithm is proposed. Since the channel estimation is always needed by the data transmission, our newly proposed algorithm can improve the synchronization performance without too much increase in the complexity. The simulations prove the effectiveness of the newly proposed algorithm.

Digital-controlled analog beamforming, which has both low cost advantage of analog beamforming and high flexibility advantage of digital beamforming, is becoming more and more attractive to wireless communication systems. Currently, the algorithm to do antenna weight training for MIMO multipath channel has very high complexity, it needs either complex MIMO channel estimation or matrix inverse, or both of them. In this paper, we propose a new two-layer iterative algorithm to train the antenna weight for the MIMO multipath channel. With this algorithm, the converged antenna weights can generate multiple beams to the directions that have strong multipaths. The simulations show the effectiveness of the newly proposed algorithm.

A distributed space-frequency (SF) coded cooperative technique is proposed for broadband wireless ad-hoc networks, where both the channels from the source node to relay nodes and from the relay nodes to the destination node are characterized by frequency-selective fading. Using an SF coding at the source node and a circular shift at each relay node, we propose a scheme that can achieve both cooperative and multi-path diversity. The pairwise error probability is analyzed and the result demonstrates that a diversity gain of min(ML1, ML2) can be achieved, where M is the number of the relay nodes, and L1and L2are the number of taps of the multipath fading channels from the source node to the relay nodes and from the relay nodes to the destination node, respectively. Furthermore, the proposed distributed SF coding can achieve full-rate transmission for any number of relay nodes. In particular, it does not need IFFT/FFT processing or decoding at each relay node, thus providing a low-complexity design of the relay nodes in broadband wireless ad-hoc networks.

A distributed space-frequency (SF) coded cooperative technique is proposed for broadband wireless ad-hoc networks. By employing an SF coding at the source node and a circular shift at each relay node, the proposed cooperative diversity technique can exploit the maximum relay diversity and multipath diversity. Pairwise error probability analysis shows that the achievable diversity gain is min (ML1, ML2), where M is the number of the relay nodes, and L1and L2are the number of taps of the multipath fading channels from the source node to the relay nodes and from the relay nodes to the destination node, respectively.

In marked contrast with the ideal error-free feedback assumption that is common in the literature, practical systems are likely to have severely bandwidth-limited, error-prone feedback channels. We consider the scenario where feedback from the receiver is used by the transmitter to select the best antenna, out of many available antennas, for data transmission. Feedback errors cause the transmitter to select an antenna different from the one signaled by the receiver. We show that optimizing the signaling assignment, which maps the antenna indices to the feedback codewords, improves performance without introducing any additional redundancy. For a system that uses error-prone feedback to transmit quadrature-phase-shift-keying-modulated data from a single antenna selected from many available spatially correlated antennas, we derive closed-form approximations for the data symbol error probability for an arbitrary number of receive antennas. We use these to systematically find the optimal signaling assignments using a low-complexity algorithm. The optimal signaling is intimately coupled to how the receiver performs selection verification, i.e., how it decodes the data signal when, due to feedback errors, it does not always know which antenna was used for data transmission. We show that ignoring feedback errors at the receiver can lead to an unacceptable performance degradation, and develop optimal and suboptimal, blind and nonblind selection-verification methods. With a small side-information overhead, nonblind verification approaches the ideal perfect selection-verification performance

In current cooperative communication schemes, to achieve cooperative diversity, synchronization between terminals is usually assumed, which may not be practical since each terminal has its own local oscillator. In this paper, based on the stack construction proposed by Hammons and El Gamal, we first construct a family of space-time trellis codes for BPSK modulation scheme that is characterized to possess the full cooperative diversity order without the synchronization assumption. We then generalize this family of the space-time trellis codes from BPSK to higher order QAM and PSK modulation schemes based on the unified construction proposed by Lu and Kumar. Some diversity product properties of space-time trellis codes are studied and simplified decoding methods are discussed. Simulation results are given to illustrate the performance of the newly proposed codes

We consider a MIMO system where error-prone feedback from the receiver is used by the transmitter to select a single optimum antenna to transmit data. Such error-prone feedback is common in the bandwidth-limited real systems, and is in marked contrast with the idealizations assumed in the selection literature. We show how the signaling assignment, which maps the antenna indices to the codewords that are fed back to indicate the index of the best transmit antenna, affects the performance of transmit antenna selection. The impact is intimately coupled with the receiver design. We derive approximate closed-form expressions for the average symbol error probability of QPSK modulated data in a spatially correlated channel, and then systematically find the optimal signaling assignment. Performance improvements are demonstrated for different antenna topologies without introducing any additional redundancy.

Using OFDM and space-time coding to achieve asynchronous cooperative diversity has been proposed by Mei et. al. and Shin et al, where orthogonal space-time block codes, Alamouti code in particular, were used at relay nodes. This proposed space-time coding can not achieve the multipath diversity when the channels from nodes to nodes are frequency selective fading. In this paper, we also consider asynchronous cooperative communications where the channels are assumed to be frequency-selective fading. We design high rate distributed space-frequency codes to achieve both full asynchronous cooperative diversity and multipath diversities.

In user cooperative communications, relay nodes are usually asynchronous. By realizing that the processing in the frequency domain is insensitive to the errors in the time domain, Mei et. al. [13] recently applied the OFDM technique to achieve the full cooperative diversity for asynchronous cooperative communications, where orthogonal space-time block codes, Alamouti code in particular, were used for relay nodes and the channels from nodes to nodes are assumed flat fading. In this paper, we also consider asynchronous cooperative communications, but the channels from nodes to nodes are assumed to be frequency-selective fading. We use high rate space-frequency codes to encode the correctly detected information symbols across some subcarriers at relay nodes to achieve both full cooperative diversity and full multipath diversity for asynchronous communications, which, we show, holds if the recent high rate space-frequency codes constructed in [15] are used.

In this letter, we study iterative demodulation and decoding methods for lattice based space-time coded systems concatenated with convolutional codes. We first apply the optimal MAP demodulation and the suboptimal linear MMSE methods to lattice based space-time demodulation and decoding. We then propose two other methods based on the idea of soft interference cancellation and in one method vector Gaussian approximation is used and in the other method scalar Gaussian approximation is used. We also present the EXIT chart analyses for the performance analyses for these iterative methods. Both theoretical and simulation results show that the performance of the vector Gaussian approximation method is the same as that of the linear MMSE method but in some cases the vector Gaussian approximation method has lower complexity. The complexity of the scalar Gaussian approximation method is the lowest among these different methods and grows linearly with the transmission rate while it still has an acceptable performance. Full diversity and full rate lattice based space-time coding is compared with the BLAST scheme and simulations show that the former has a better performance than the later even with the proposed suboptimal iterative demodulation and decoding methods.

In this paper, we consider a coded ultra-wide band (UWB) impulse radio system. At the receiver, to estimate multipath delays and amplitudes, the soft information output from a soft-input soft-output (SISO) decoder is used to assist the estimation. The SISO decoder and the channel estimator work in an iterative way and no training sequence is used in the estimation. Compared with the non data-aided channel estimation without using the soft information output from the SISO decoder, the performance of this iterative channel estimation is better. We also explain this iterative method by using the expectation maximization (EM) algorithm, which can be used to prove the convergence of this iterative method.

In current cooperative communication schemes, to achieve cooperative diversity, synchronization between terminals is usually assumed which may no be practical since each terminal has its own local oscillator. In this paper, based on the stack construction proposed by Hammons and El Gamal, we first construct a family of space-time trellis codes from BPSK to higher order QAM and PSK modulation schemes based on the unified construction proposed by Lu and Kumar Some diversity product properties of space-time trellis codes are studied and simplified decoding methods are discussed. Simulation results are given to illustrate the performance of the newly proposed codes.

In current cooperative communication schemes, to achieve cooperative diversity, synchronization between terminals is usually assumed, which may not be practical since each terminal has its own local oscillator. In this paper, based on the stack construction proposed by Hammons and El Gamal, we first construct a family of space-time trellis codes for BPSK modulation scheme that is characterized to possess the full cooperative diversity order without the synchronization assumption. We then generalize this family of space-time trellis codes from BPSK to higher order QAM and PSK modulation schemes based on the unified construction proposed by Lu and Kumar. And also, simplified decoding methods are discussed. Simulation results are given to illustrate the performance of the newly proposed codes.

The mapping from binary sequences to space-time matrices is an important issue in designing space-time coded multiple-antenna systems where different mapping schemes may lead to different bit-error-rate (BER) performance. In this letter, we use a binary switching algorithm to investigate the mappings of bit-interleaved coded modulation with iterative demodulation/decoding (BICM-ID) for space-time modulation. Both coherent and noncoherent space-time modulation in the cases of no a priori information and perfect a priori information of the coded bits for the demodulator are considered.

Signal-space diversity is a power- and bandwidth-efficient diversity technique. To exploit the signal-space diversity, joint maximum-likelihood (ML) detection at the receiver is usually needed, where the complexity grows exponentially with the dimension of the lattice. In this letter, we propose a serial concatenated scheme and two simple iterative methods to exploit the signal-space diversity. The simple iterative methods are based on the idea of soft interference cancellation. The first iterative method is based on a vector Gaussian approximation, while the second one is based on a scalar Gaussian approximation. The complexity of the first iterative method grows cubically with the dimension of the lattice, and the simulations show that its performance approaches that of the optimal maximum a posteriori detection method. The complexity of the second iterative method grows linearly with the dimension of the lattice, and the simulations show that when the dimension of the lattice N=32, at bit-error rate =10/sup -5/, the performance gap between the Rayleigh fading channel and the Gaussian channel is only 0.3 dB.