We address the problem of synthesis and generation of faces from edgemaps, motivated by extreme low bit-rate facial compression and the need for robust source-channel coding over noisy channels. Three approaches for image reconstruction are proposed. In the first, a deep learning-based encoder-decoder creates a latent space representation of the original image. An Edgemap-to-Latent Mapper (ELM) network maps the input edgemap to this latent space, with the final image reconstructed using a pre-trained compressive decoder. The second approach retrains the compressive decoder to reconstruct images from the ELM network’s output. The third approach jointly trains the ELM network and decoder, enabling direct reconstruction from the edgemap. This end-to-end framework achieves reasonable reconstruction fidelity. We also examine the impact of additive channel noise on edgemap transmission under low SNR conditions, demonstrating that even with significant noise, a DNN-based joint denoiser and edgemap decoder can reconstruct images. At extremely low SNRs, where edgemaps are highly corrupted, the network also exhibits generative capabilities, producing plausible images.

In this paper, we consider an Orthogonal Frequency Division Multiplexing (OFDM) system with a periodic pilot insertion pattern on the Time-Frequency (TF) grid. We compute Ambiguity functions at the receiver, which are two-dimensional (2D) delay-Doppler (DD) domain correlations of transmitted and received signals. Using these ambiguity functions, we propose a channel estimation technique for sparse DD channels. The TF channel estimate can be obtained by directly computing the 2D Fourier Transform of the ambiguity functions. While this approach gives better performance than conventional frequency-domain channel estimation and interpolation for OFDM, it does not exploit the sparsity in the DD domain. Our proposed approach uses the inherent sparsity of the ambiguity functions and estimates the sparse DD channel using Orthogonal Matching Pursuit. The obtained sparse DD channel is then used to calculate the TF channel. Through simulations, we study both the direct approach and our proposed sparsity-based approach and show improved channel estimation and error performance using the DD domain approaches.

In the context of cellular networks, users located at the periphery of cells are particularly vulnerable to substantial interference from neighbouring cells, which can be represented as a two-user interference channel. This study introduces two highly effective methodologies, namely TwinNet and SiameseNet, using autoencoders, tailored for the design of encoders and decoders for block transmission and detection in interference-limited environments. The findings unambiguously illustrate that the developed models are capable of leveraging the interference structure to outperform traditional methods reliant on complete orthogonality. While it is recognized that systems employing coordinated transmissions and independent detection can offer greater capacity, the specific gains of data-driven models have not been thoroughly quantified or elucidated. This paper conducts an analysis to demonstrate the quantifiable advantages of such models in particular scenarios. Additionally, a comprehensive examination of the characteristics of codewords generated by these models is provided to offer a more intuitive comprehension of how these models achieve superior performance.

Orthogonal Time Frequency Space (OTFS) receivers perform joint equalization of all the symbols in an OTFS frame. We consider delay-time domain Maximal Ratio Combining (MRC) equalization for OTFS systems with receiver oversampling. The associated impairments are Inter-Frame Interference (IFI) and Inter-Block Interference (IBI). The existing Zero-Padding (ZP) approach avoids IFI and IBI but requires high overhead. We propose a Reduced Cyclic Prefix (RCP) and Reduced Cyclic Suffix (RCS) insertion method on the transmitter side that achieves high throughputs. Our approach avoids IFI, and we modify the existing equalizer designed for ZP systems to handle IBI without a significant increase in complexity. We present the Bit Error Rate (BER) simulation results for different oversampling factors and explain the diminishing returns.

The massive multiple-input multiple-output (MIMO) full-duplex two-way relaying (FD-TWR) literature has extensively investigated power scaling for rate guarantees by considering a fixed number of users. We investigate the pairwise error probability (PEP) and the per-user rate of a FD-TWR with N, relay antennas that employs maximal ratio combining/transmission to enable two-way communication between K FD users. We propose novel relay and user powers scalings, with both N, and K tending to infinity, and show that the PEP of each user converges almost surely to its AWGN counterpart. These power scalings are different from the existing ones, which are derived by fixing K and by assuming that only N, tends to large values. We show that the analysis developed herein applies to both Gaussian and non-Gaussian complex channels with finite number of moments. We numerically show that when both K and N, increase concurrently to large values, the proposed power scaling schemes not only have better per-user PEP and rate than the existing schemes, but they are also robust to the FD self loop-interference power.

Faster-Than-Nyquist (FTN) Signaling is a non-orthogonal transmission scheme which violates the Nyquist zero-ISI criterion providing higher throughput and better spectral efficiency than a Nyquist transmission scheme. This comes with a cost of higher transceiver complexity. In this paper, we focus on understanding pulse shapes and their inter-symbol-interference (ISI) and show that, under certain conditions on pulse shapes and τ (time acceleration factor), the ISI can be avoided completely with the help of precoding. This leads to a symbol-by-symbol detection. Further, we extend this idea to Orthogonal Frequency Division Multiplexing (OFDM) FTN systems and show that, under certain conditions, the average performance of OFDM system reaches that of a Nyquist system. Finally, simulation results of the performance of precoded and non-precoded single carrier and OFDM FTN systems are compared to a Nyquist system.

The maturing massive multiple-input multiple-output (MIMO) literature has provided asymptotic limits for the rate and energy efficiency (EE) of maximal ratio combining/ maximal ratio transmission (MRC-MRT) relaying on two-way relays (TWRs) using the amplify-and-forward (AF) principle. Most of these studies consider time-division duplexing and a fixed number of users. To fill the gap in the literature, we analyze the MRC-MRT precoder performance of an N-antenna AF massive MIMO TWR, which operates in a frequency-division duplex mode to enable two-way communication between 2M = [Nα] single-antenna users, with α ∈ [0, 1), divided equally into two groups of M users. We assume that the relay has realistic imperfect uplink channel state information (CSI), and that quantized downlink CSI is fed back by the users relying on B ≥ 1 bits per-user per relay antenna. We prove that for such a system with α ∈ [0, 1), the MRC-MRT precoder asymptotically cancels the multi-user interference (MUI) when the supremum and infimum of large-scale fading parameters are strictly nonzero and finite, respectively. Furthermore, its per-user pairwise error probability converges to that of an equivalent AWGN channel, as both N and the number of users 2M = [Nα] tend to infinity, with a relay power scaling of Pr= (2MEr/N) and Erbeing a constant. We also derive upper bounds for both the per-user rate and EE. We analytically show that the quantized MRC-MRT precoder requires as few as B = 2 bits to yield a BER, EE, and per-user rate close to the respective unquantized counterparts. Finally, we show that the analysis developed herein to derive a bound on α for MUI cancellation is applicable both to Gaussian as well as to any arbitrary non-Gaussian complex channels.

We design a precoder for non-concurrent two-way relaying (ncTWR) where a base station (BS) serves a transmitonly user equipment (TUE) in the uplink and a receive-only user equipment (RUE) in the downlink. The RUE experiences backpropagating interference (B!). The proposed precoder is designed such that it not only cancels the B! experienced by the RUE but more importantly, enables receive data decoding with high diversity. The high diversity precoder is designed by deriving the closed-form pairwise error probability (PEP) expressions, and by optimizing the precoder elements to minimize the PEP. We analytically show that with Nr-antenna relay, Nb-antenna BS, and Nu-antenna TUE and RUE, both BS and RUE decode their respective data with a diversity order of min(Nu2, (Nr- Nu)Nb) at high receive signal-to-noise ratio. We also numerically show that the proposed design has lower bit error rate than the existing state-of-the-art ncTWR designs.

We investigate a limited feedback precoder based on symbol pairwise error probability (PEP) for a block-faded K×ntdownlink multiple-input multiple-output (MIMO) channel. In the considered system, K = ⌊ntα⌋ single-antenna users feedback quantized channel state information to the nt-antenna transmitter using B bits per-transmit-antenna per user. We analytically show that for αt→ ∞, both symbol PEP and achievable rate of each of the K downlink users almost surely converge to the symbol PEP and achievable rate of K parallel additive white Gaussian noise (AWGN) channels, respectively. We show that the encoding complexity of the precoder is O(ntK). We also show that if channel coefficients estimated by the user are corrupted by AWGN noise, the symbol PEP and achievable rate of each user almost surely converge to the symbol PEP and achievable rate in a scaled AWGN channel with B > 1 and nt→ ∞. For correlated channels, we derive a condition, which enables the proposed precoder almost surely to cancel multi-user interference for large ntvalues. Finally, we numerically compare the bit error rate, encoding complexity, and per-user achievable rate of the proposed scheme with the existing designs.

In this paper, we propose two-hop user-deployable indoor relays for CDMA systems with signal bandwidth 5 MHz or less. We assume concurrent relaying wherein macro base stations (BS) are allowed to transmit when a relay is forwarding. The two-hop indoor relay is studied using a generalized model for outdoor-to-indoor communication in a frequency-selective channel in the presence of co-channel interference. We propose an equalize-and-forward relay that employs multiple-antenna interference-aware equalization before forwarding. We present a unified solution to minimize interference (due to concurrent transmissions) at the relay-served destination and the BS-served destination, using power control and exploiting spatial reuse. A systematic method to enable indoor relays with the proposed interference management solutions is presented that is applicable to existing CDMA systems, without additional requirements in Access Terminal (AT)/User Equipment. Simulation results in the context of High Rate Packet Data (HRPD) CDMA network shows that the ATs served through the indoor relays experience throughput improvement by at least a factor of two, whereas the ATs served by the BS experienced a gain of around 6% in spite of concurrent relaying. The results also show an overall system throughput gain upto 30% when compared to the existing network with only single hops.

We address the problem of designing a communication system for deploying under disaster situations where the pre-disaster communication infrastructure has been uprooted. The disaster communication system is required to aid in rescue effort by providing text, voice, and video communication links between the field worker and the operation center. It is also required to provide some communication services to the affected citizens so that the overwhelming demands made on the communication infrastructure during the times of disaster are mitigated to some extent. The other highlights of the design are its low-cost nature by exploiting the economies of scale and the fact that basic mobile phones with text-only features are widely available in developing countries.

Performance of spatial multiplexing multiple-input multiple-output (MIMO) wireless systems can be improved with channel state information (CSI) at both ends of the link. This paper proposes a new linear diagonal MIMO transceiver, referred to as co-ordinate interleaved spatial multiplexing (CISM). With CSI at transmitter and receiver, CISM diagonalizes the MIMO channel and interleaves the co-ordinates of the input symbols (from rotated QAM constellations) transmitted over different eigenmodes. The analytical and simulation results show that with co-ordinate interleaving across two eigenmodes, the diversity gain of the data stream transmitted over the weaker eigenmode becomes equal to that of the data transmitted on the stronger eigenmode, resulting in a significant improvement in the overall diversity. The diversity-multiplexing tradeoff (DMT) is analyzed for CISM and is shown that it achieves higher diversity gain at all positive multiplexing gains compared to existing diagonal transceivers. Over rank n MIMO channels, with input symbols from rotated n-dimensional constellations, the DMT of CISM is a straight line connecting the endpoints (0,NtNr) and (min{Nt,Nr}, 0), where Nt, and Nrare the number of transmit and receive antennas, respectively.

Vertical layering with successive interference cancellation (SIC) at the receiver (e.g. V-BLAST) is a popular spatial multiplexing scheme that achieves high data rates over multi-antenna wireless channels. However, SIC suffers from low diversity gains and recently, with limited feedback, a diversity optimal SIC receiver has been proposed based on greedy QR decomposition (GQR-SIC). But the low diversity gain of the first decoded layer degrades the overall diversity gain. In this paper, we present a new diagonally layered spatial multiplexing scheme that employs GQR-SIC receiver but achieves significantly higher diversity gains. In the proposed scheme, the diversity gain of a weak layer is improved through interleaving the co-ordinates of the symbols transmitted over the weak layer and a strong layer. In this work, we focus on spatial multiplexing with only two layers and analyze the diversity multiplexing tradeoff to show that the proposed scheme achieves maximum diversity gain. When compared with other diagonally layered schemes, the proposed scheme has low decoding complexity.

When channel state information is available at the transmitter, we can diagonalize a MIMO channel with SVD transceivers. However, outage of the weaker eigenchannels limits the performance of such a transceiver. In the current work, we propose and analytically characterize co-ordinate interleaving of multi-dimensional symbols over sets of eigenchannels. We show that in our method, the diversity multiplexing trade-off (DMT) is determined by the strongest eigenchannel in each set. We also calculate the optimal (DMT) for different possible sets and constellation dimensions with rate allocation among the sets.

Spatial multiplexing over multiple-input multiple-output (MIMO) channels significantly improves the data rates over wireless channels. With channel knowledge at both ends of the link, many spatial multiplexing schemes have been proposed in literature, showing improved performance with respect to some performance measure. Co-ordinate interleaved spatial multiplexing (CISM) (Srinivas et al., 2006) is one such scheme that has been shown to achieve improved error rate performance. In this paper, we analyze CISM from diversity-multiplexing gain tradeoff (DMT) perspective. We focus on CISM over MIMO channels having only two degrees of freedom and show that the DMT of CISM is a straight line connecting the two extreme points (0, 2 m) and (2, 0) where m = max{Nt, Nr}. This result shows that CISM outperforms many other schemes that require same amount of feedback.

Spatial multiplexing (SM) over multiple-input multiple-output wireless channels provides significant capacity gains. In a SM scheme, the eigenmode having the least signal-to-noise ratio (SNR), degrades the overall error rate performance. In this paper, we propose co-ordinate interleaved spatial multiplexing that maximizes the minimum SNR over all eigenmodes. This linearly decodable SM scheme needs the knowledge of the right singular vectors of the channel at the transmitter, and the singular values and left singular vectors at the receiver. We derive the SNR expressions for the proposed scheme and compare its performance with other closed-loop schemes using computer simulations

This paper describes the physical layer for a high penetration paging (HPP) mode, that is a highly integrated adjoint to a cellular system. Through the use of very low rate coding and periodically inserted synchronization bursts, the HPP mode enables a cellular user to locate and maintain a minimal communication link with a cellular network at 15 to 20 dB below the normal operating signal to noise ratio (SNR). In this paper, the ANSI-136 air interface standard is used as the underlying cellular system. The same concept can be extended to other systems as well.

Doppler spread estimation in digital mobile radio systems is described. An optimal approach (in the sense of maximum likelihood estimation) for Doppler spread estimation is derived. Suboptimal approaches, with trading off in performance and complexity, are developed from the optimal algorithm. In practical systems, the Doppler spread must be estimated in the presence of frequency offsets between the transmitter and the receiver. Hence, an algorithm that decouples Doppler estimation from automatic frequency compensation (AFC) is also presented. The proposed algorithms rely on channel estimation over the known fields of a TDMA burst. Such algorithms can be applied easily to systems with pilot symbols or a pilot channel.

In this paper, efficient and practical approaches for Doppler spread estimation in wireless mobile radio systems are described. A hypothesis-testing approach, given the channel autocorrelation estimate, is utilized for Doppler spread estimation. The channel autocorrelation is estimated slot-by-slot using the knowledge of the channel estimates over the known fields of a TDMA burst, and averaged over several slots to reduce the effect of noise. In practical systems, the Doppler spread must be estimated in the presence of frequency offsets between the transmitter and the receiver. Hence, an algorithm that decouples Doppler spread estimation from automatic frequency compensation (AFC) is also presented. In addition to the mean estimation error performance, the convergence and the tracking ability of the algorithms are evaluated via simulation using ANSI 136 Rev-B modulation and signal transmission format.

The use of the "software radio" concept in cellular applications is a topic of widespread interest. Two key issues in the implementation of software radios are the development of optimal receivers that require the minimum number of bits in the wide-band analog-to-digital converter (ADC) and efficient channelizers that extract individual channels from the digitized wide-band signal. In this paper, both of these issues are studied in detail for cellular base stations. A computationally efficient wide-band channelizer is presented. This channelizer is closely related to the discrete Fourier transform filter bank used in transmultiplexers. It is shown that the complexity of the proposed channelizer is significantly less (2-50/spl times/) than the complexity of conventional channelizers. An optimal receiver that explicitly takes into account the effect of the quantization noise of the wide-band ADC is also derived. The analysis of the ADC noise provides guidelines for specifying wide-band ADC for use in cellular applications. The development of the channelizer and the optimal receiver yield important insights into the implementation of cellular software radios. All of the key results of this paper are applied to a detailed example based on the Digital Advanced Mobile Phone System (D-AMPS, IS-54/IS-136) cellular standard. The bit-error rate (BER) performance simulations of a D-AMPS wide band receiver is presented as a part of this example.

For cellular software radio receivers, this paper presents a computationally efficient algorithm for extracting individual radio channels from the output of the wideband A/D converter. In a software radio, the extraction of individual channels from the output of the wideband A/D converter is by far the most computationally demanding task; hence, it is very important to devise computationally efficient algorithms for this task. Our algorithm is obtained by modifying the DFT filter bank structure that is well known in the multi-rate signal processing literature. We show that the complexity of the proposed algorithm is significantly less (2X-50X) than the complexity of the conventional channelizers.

For cellular software radio receivers, this paper presents a computationally efficient algorithm for extracting individual radio channels from the output of the wideband A/D converter. In a software radio, the extraction of individual channels from the output of the wide band A/D converter is by far the most computationally demanding task; hence, it is very important to devise computationally efficient algorithms for this task. Our algorithm is obtained by modifying the DFT filter bank structure that is well known in the multi-rate signal processing literature (Scheuermann and Gockler, 1981; Vaidyanathan, 1990). We show that the complexity of the proposed algorithm is significantly less (2/spl times/-50/spl times/) than the complexity of the conventional channelizers.

Achievable spectral efficiency in land-mobile radio (LMR) systems is limited by adjacent-channel interference (ACI) specifications. Spectral efficiency can be improved by using digital modulation schemes. With linear modulation schemes, intermodulation products produced by non-linear power amplifiers lead to increased ACI. In this paper, a new family of linear modulation schemes called envelope-constrained modulation schemes is introduced to improve the spectral efficiency of LMR systems using quasilinear power amplifiers. These modulation schemes exhibit low envelope variations and thus ease the linearity requirements on the power amplifier. Also, power savings can be realized by incorporating coded modulation. An analytical evaluation of the performance is performed to illustrate the gain achievable using the new modulation schemes.

Set partitioned trellis codes and multi-level block codes can be used to increase the information rate of the American digital cellular system (IS-54) from 1.22 bits/symbol to 1.75-2 bits/symbol. This increase in information rate can be used to enhance speech quality or increase system capacity. The bit error rate performance of several block coded modulation schemes is studied via simulation for flat and dispersive Rayleigh fading channels using the IS-54 base-mobile frame format. Coherent and differential detection, and equalization are studied considering finite interleaving and channel estimation via interpolation.>

The spectral efficiency in land mobile radio (LMR) systems is limited by stringent adjacent-channel interference protection ratio (ACIPR) specifications of at least 60 dB. High spectral efficiency can be achieved using digital modulation techniques. Linear modulation schemes exhibit high spectral efficiencies but require linear power amplifiers. Power amplifiers typically used in LMR are nonlinear which degrade the ACIPR for conventional linear modulation schemes. The paper presents simulated performance results of a trellis-coded linear modulation scheme that is designed to be less sensitive to the nonlinearity exhibited by the power amplifier. Simulation results indicate that an ACIPR of 60 dB can be achieved. An E/sub b//N/sub 0/ of 14 dB is required to achieve a bit error rate of 3% with 3 users per 25 kHz channel, with an average data rate of 9.6 kbits/second/user.>

In mobile communications, the receiver must mitigate severe signal impairments arising from multipath and Rayleigh fading. Multipath produces intersymbol interference (ISI), necessitating the use of an equalizer in the receiver. The performance of the equalizer can be significantly improved using antenna diversity, thereby allowing the use of lower C/I values. This C/I reduction offers potential for capacity enhancement. This paper describes the use of pre-detection antenna diversity in conjunction with a MLSE receiver for the U.S. Digital Cellular (USDC) system. It is shown that the receiver meets all the USDC (IS-54) performance specifications with a significant margin and has a computational complexity of less than 4 MIPS.>

The authors present a derivation of the necessary and sufficient condition on the polyphase components of a linear-phase prototype (length N=2 mM) such that the polyphase component matrix of the filter bank is lossless. This in turn ensures that the modulated filter bank satisfies the PR (perfect reconstruction) property. An efficient design procedure (which involves fewer parameters to be optimized than other approaches) is presented. By this approach, FIR (finite-impulse-response) PR filter banks can be designed for an arbitrary number of channels. Moreover, since both the analysis and synthesis filter banks are obtained by modulation, they can be implemented very efficiently (using the discrete cosine transform). The design procedure is outlined and a design example is presented.>