Zhenning Shi, Yijia Zhu, Yi Xie, Junhan Shi, Guorui Xie, Haotian Zhang, Yong Jiang, Congcong Miao, Qing Li. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

Network Intrusion Detection Systems (NIDS) are critical for web security by identifying and blocking malicious traffic. In-network NIDS leverage programmable switches for high-speed traffic processing. However, they are unable to reconcile the fine-grained classification of known classes and the identification of unseen attacks. Moreover, they lack support for incremental updates. In this paper, we propose Helios, an in-network malicious traffic detection system, for continual adaptation in attack-incremental scenarios. First, we design a novel Supervised Mixture Prototypical Learning (SMPL) method combined with clustering initialization to learn prototypes that encapsulate the knowledge, based on the weighted infinity norm distance. SMPL enables known class classification and unseen attack identification through similarity comparison between prototypes and samples. Then, we design boundary calibration and overlap refinement to transform learned prototypes into priority-guided matching rules, ensuring precise and efficient in-network deployment. Additionally, Helios supports incremental prototype learning and rule updates, achieving low-cost hardware reconfiguration. We implement Helios on a Tofino switch and evaluation on three datasets shows that Helios achieves superior performance in classifying known classes (92%+ in ACC and F1) as well as identifying unseen attacks (62% - 98% in TPR). Helios has also reduced resource consumption and reconfiguration time, demonstrating its scalability and efficiency for real-world deployment.

Recently, research on denoising diffusion models has expanded its application to the field of image restoration. Traditional diffusion-based image restoration methods utilize degraded images as conditional input to effectively guide the reverse generation process, without modifying the original denoising diffusion process. However, since the degraded images already include low-frequency information, starting from Gaussian white noise will result in increased sampling steps. We propose Resfusion, a general framework that incorporates the residual term into the diffusion forward process, starting the reverse process directly from the noisy degraded images. The form of our inference process is consistent with the DDPM. We introduced a weighted residual noise, named resnoise, as the prediction target and explicitly provide the quantitative relationship between the residual term and the noise term in resnoise. By leveraging a smooth equivalence transformation, Resfusion determine the optimal acceleration step and maintains the integrity of existing noise schedules, unifying the training and inference processes. The experimental results demonstrate that Resfusion exhibits competitive performance on ISTD dataset, LOL dataset and Raindrop dataset with only five sampling steps. Furthermore, Resfusion can be easily applied to image generation and emerges with strong versatility. Our code and model are available at https://github.com/nkicsl/Resfusion.

Neural networks (NNs) are widely used in classification-based networking analysis to help traffic transmission and system security. However, there are heterogeneous network devices (e.g., switches and routers) in a network. Manually customizing NNs with specific device requirements (e.g., max allowed running latency) can be time-consuming and labor-intensive. Furthermore, the diverse data characteristics of different networking classification tasks add to the burden of NN customization. This paper introduces Loong, a neural architecture search (NAS) based system that automatically generates NNs for various networking tasks and devices. Loong includes a neural operation embedding module, which embeds candidate neural operations into the layer to be designed. Then, the layer-wise training is used to generate a task-specific NN layer by layer. This layer-wise scheme simultaneously trains and selects candidate neural operations using gradient feedback. Finally, only the important operations are selected to form the layer, maximizing accuracy. By incorporating multiple objectives, including deployment memory and running latency of devices, into the training and selection of NNs, Loong is able to customize NNs for heterogeneous network devices. Experiments show that Loong's NNs outperform 13 manual-designed and NAS-based NNs, with a 4.11% improvement in F1-score. Additionally, Loong's NNs achieve faster (7.92X) speeds on commodity devices.

Interference coordination is a critical enabler of heterogeneous networks. In this paper we present a scheduling cooperation scheme that coordinates the resource allocation and user assignment across layers. The proposed scheme focuses on mitigating the interference generated by macro-associated users in the proximity of small cells by exchanging the information of user assignment on the severely-interfered spectrum. Users that are most resilient to the macro-introduced interference are selected for transmission on the shared channel. In legacy networks the coordination scheme is subject to backhaul latency restriction on the inter-node signalling. We then present an enhanced scheduler based on predicted channel conditions to improve the system performance in the presence of non-ideal backhauls.

Heterogeneous deployment comprising overlays macrocell and small cells is a promising solution to enhance both system capacity and coverage. Nevertheless, the performance of heterogeneous networks uplink is severely degraded due by cross-layer interference, especially the interference that small cell suffers from the proximate macro UEs (MUEs). In this paper, we propose a Cross-Tier Interference-Aware (CTIA) scheduling scheme which exploits scheduling coordination and small cell spacial degree of freedom to mitigate the uplink interference. The system level simulations show that substantial throughput gains can be achieved with the proposed CTIA scheme and practical receivers. It is also suggested to properly balance between the capacity improvement achieved by the proposed method and the backhaul overhead consumption, which can be implemented by adapting the cooperating set size with a system-defined parameter.

In this paper, we investigate efficient beamforming schemes for downlink broadcast channels of MIMO systems. Optimal solution to this problem is known to be infeasible due to the prohibitive complexity and recent research focuses on suboptimal approaches with reduced complexity. We first present a progressive beamforming method that maximizes the net gain in sum throughput for newly added users and achieves high sum throughput in MU-MIMO channels. Nevertheless, there is severe asymmetry in the performance between data streams: data streams scheduled at an earlier stage are better protected compared to their counterparts configured in later stages. We then show that inter-stream asymmetry can be substantially alleviated by updating beamforming vectors over iterations. An adaptive algorithm is developed to enhance the efficiency of iterative processing and is shown to substantially improve the system performance with manageable complexity.

Capacity and Coverage are the two major concerns for the wireless operators to deploy cellular networks. This paper investigates the uplink capacity and coverage of relay-assisted UMTS network when multiuser detection techniques are employed at both macrocell base station (BS) and relay station (RS). A hierarchical geometric system model is used where the fixed relay stations are deployed within the macrocell. We are particularly interested in the cases where macro BSs and RSs have different capabilities in mitigating intracell and intercell interferences. Simulation results show that the capacity can be tripled and the cell range can be extended when MUD can be employed at base stations and relay stations. Furthermore, we studied the practical scenario where the users in the relay region are partially served by the relay station. This case is meaningful when there are not enough relay resources or users are restricted due to limited handover capability. We then proposed a partial relay system to enhance the cell range while maintaining a reasonable cell load.

Since the Quality of Experience (QoE) metric directly reflects the subjective experiences of users, the resource allocation techniques with the objective to improve QoE instead of conventional metrics are efficient in resource utilization. This kind of techniques can avoid wasting resources in providing perfect objective metric which may have little or even no effect on user experience. This paper presents a QoE-aware power allocation algorithm in multiuser Orthogonal Frequency Division Multiplexing (OFDM) systems. The power allocation problem is formulated to maximize the overall QoE subject to the total transmit power constraint. The original model is solved by means of transformation and approximation, since it is difficult to develop a closed-form solution. Simulation results show that the proposed algorithm can enhance the performance in terms of both the overall QoE and single user QoE in comparison with the classical water-filling algorithm.

Multicell collaboration is considered for power allocation in OFDMA downlink to mitigate the intercell interference. We propose a coordination approach to maximize the sum utility of a cluster of co-channel users by properly configuring user transmit power in local cells. In particular, we employ the well-known α-PF metric as the utility function to accounts for the fairness between users in different cells. It is arguably shown that optimization of fairness-oriented utilities in the multicell scenario through intercell coordination effectively mitigates the interference and improves signal transmission. Our work sheds new lights on intercell interference coordination in multicarrier downlink, and leads to a distributed framework for multicell resource coordination.

In the multi-cell environment, the approaches that are suitable for the single cell scenario may cause excessive interference to other cells and become suboptimal. In this paper, we present a number of receiver-centric power allocation schemes that take into account the detrimental impact of allocated resources to the neighboring cells, and use it as a guideline in determining the appropriate power levels over the spectrum. Together with a user scheduler with proportional fairness, the proposed methods demonstrate a significant improvement in the cell edge performance at barely no compromise in the average user throughput.

It is well-known that wireless scheduling algorithm could exploit multi-user diversity to enhance the network capacity. With multiple transmit/receive antennas, there are additional degrees of freedom which could deliver either spatial multiplexing gain and/or spatial diversity gain. With cross layer scheduling, there is also multi-user selection diversity which contributes to both network capacity and coverage. Due to the huge complexity involved for the optimal solution and big performance gap caused by greedy schemes, in this paper, we proposed a series of genetic algorithm based scheduling schemes with different chromosome code designs for both single- and multi-cell system resource allocation. For single-cell scheduling, we mainly focused on uplink multi-user scheduling with successive interference cancellation at receivers, and joint spatial-frequency scheduling. While for multi-cell scheduling, we proposed a centralized genetic algorithm based scheduling solution that could effectively increase the multi-cell cooperation gain. Simulation results show that the proposed algorithms could fill in most of the performance gap at reasonable complexity compared with optimal solution and greedy scheduling.

In this letter, we propose a successive multiuser detector (SMUD) for contention based OFDMA ranging channel compliant to the IEEE 802.16 (WiMAX) standard. A ranging channel consists of a set of subcarriers in specific time slots shared by multiple users, so the multiple access interference (MAI) limits the performance of ranging detectors. Different from existing methods that treat the MAI as noise, the proposed SMUD successively detects the channel paths of active ranging signals and cancels their interference for further detection. This approach significantly suppresses the MAI and improves both user detection and parameter estimation performance.

In orthogonal frequency-division multiplexing (OFDM) systems, the integer part of carrier frequency offset (CFO) as a multiple of subcarrier spacing needs to be estimated from the frequency-domain signal. The performance of the autocorrelation based method is quite poor, while the channel knowledge based method is too complex for hardware real-time implementation. To provide improved tradeoffs between the performance and complexity, we propose a new hybrid algorithm that filters away unlikely integer CFO candidates with the autocorrelation method and determines the most likely estimate with channel knowledge. The analytical performance of the proposed hybrid method agrees with the simulation results and both demonstrate the cost-effectiveness of the method even in challenging wireless channel conditions.

This paper studies fractional carrier frequency offset (CFO) estimation for orthogonal frequency-division multiplexing (OFDM) systems. In the IEEE 802.16 (WiMAX) standard, a training symbol structure with three highly correlated but not identical segments is specified. The existing CFO estimation methods require the segments in the training symbol to be identical, so cannot work for the generalized training structure. In this paper, we proposed a universal fractional CFO estimator to solve the problem. Both the analytical and numerical results are presented to confirm the performance of the proposed method.

In this paper, we develop a rate allocation scheme for interleave-division multiple-access (IDMA) systems with iterative decoding. We use a fully-analytical approach to predict the performance of the scheme and propose a modified linear programming method to design the rate profile. Numerical results show that with just repetition coding and rate allocation, the performance of the scheme is only about 5 dB away from the capacity for a wide range of SNR, provided that the number of users is sufficiently large. Compared with power allocation schemes for IDMA, the proposed rate allocation scheme achieves a similar performance at a moderate spectral efficiency and the requirement of sophisticated power amplifiers can be relaxed.

To alleviate the high peak-to-average power ratio (PAPR), high complexity in user terminal and sensitivity to carrier frequency offset (CFO) problems in current OFDMA systems, a Quadrature OFDM (Q-OFDMA) system has been recently proposed in the single-input single-output environment. In this paper we investigate the realization of multi-input multioutput (MIMO) diversity- and capacity- oriented methods in Q-OFDMA systems. We construct an Alamouti-like spacetime block code (STBC) for Q-OFDMA signal symbols, and investigate the ZF and MMSE equalizers to realize the simple detection for spatial multiplexing (SM) Q-OFDMA systems.

In current OFDMA systems three major problems arise due to the large number of subcarriers, including high peak to average power ratio (PAPR), sensitivity to carrier frequency offset (CFO), and high complexity in users terminals. In this paper, based on an innovative concept of layered FFT structure, we propose novel quadrature OFDMA (Q-OFDMA) systems which can overcome these problems. In particular, the proposed systems can achieve the same guard-interval overhead and same bandwidth occupation to conventional OFDMA systems, while with reduced PAPR and improved CFO robustness and frequency diversity. Q-OFDMA systems also promise low complexity in downlink receivers. Parameter configuration is investigated for both predefined and adaptive users data rates. Theoretical comparison of bit error rate (BER) performance between Q-OFDMA and OFDMA is conducted, and validated by simulation results. It is shown that Q-OFDMA systems could achieve better performance than OFDMA when signal to noise ratio (SNR) is above a threshold depending on the channel condition, and advanced equalizers, such as minimum mean square error equalizer, can significantly decrease this threshold due to the large frequency diversity in Q-OFDMA systems.

In orthogonal frequency division multiplexing (OFDM) systems, coarse frame timing can be acquired from one or more training symbols preceding every OFDM burst. The existing literature studied the case where there was only one training symbol consisting of identical segments. We generalize the timing synchronization methods to take advantage of multiple training symbols and only require the segments to be highly correlated but not necessarily identical. We construct a series of component timing metrics, one for each pair of the highly correlated segments, and combine them linearly to minimize the false alarm probability while keeping the asymptotic missed detection probability to the same level as other techniques. The OFDM data symbols in a downlink burst can have different power levels to reach the users at different distances. We take that into account and yield more realistic results than those in existing literature which only considered the equal power case. The performance of the proposed method is analyzed in three scenarios generalized from the IEEE 802.11 and IEEE 802.16 standards. Numerical results are presented to confirm the robustness of the proposed method in various channel conditions.

In this letter, we develop a rate allocation scheme for interleave-division multiple-access (IDMA) systems with iterative decoding. We use a fully-analytical approach to predict the performance of the scheme. Then, we propose a modified linear programming method to find the best rate profile for the scheme. Numerical results show that with just repetition coding and optimal rate allocation, the performance of the scheme is only about 5 dB away from the capacity for a wide range of SNR, provided that the number of users is sufficiently large. Compared with power allocation schemes for IDMA, the proposed rate allocation scheme achieves a similar performance at a moderate spectral efficiency and the requirement of sophisticated power amplifiers can be relaxed.

To alleviate the high peak-to-average power ratio (PAPR) and sensitivity to carrier frequency offset (CFO) problems, an asymmetric OFDM (A-OFDM) system has been recently proposed. In this paper, the theoretical bit error rate (BER) analysis of the A-OFDM systems is presented. Various modulation methods, i.e., BPSK and M-ray QAM, and equalization techniques, i.e., zero forcing (ZF) and minimum mean square error (MMSE) criteria are considered. Our analysis shows that the BER performance of the A-OFDM system bridges that of conventional OFDM and single carrier systems. Certain advanced equalizers, such as MMSE equalizer, are found to significantly improve the BER performance of the A-OFDM systems. The analysis approach is generally applicable to most modulation schemes and can be extended to any preceded OFDM systems.

Acquisition of the timing in an IDMA system must take place before signal detection and decoding are performed. Acquisition in the presence of severe multiple-access interference with time varying codes makes the task even more difficult. Inefficient designs cause a large number of false alarms and/or missed detections. This paper applies a powerful acquisition technique to IDMA systems in the uplink. Under high multiple access interference conditions conventional acquisition techniques used for DS-CDMA systems simply fail. The proposed method utilises soft data from the iterative IDMA receiver to effectively cancel interference from the co-channel users. Analytical performance in terms of the number of users, spreading gain, performance of the IDMA receiver, and noise variance is developed, together with simulation results.

Quadrature OFDMA (Q-OFDMA) systems have been recently proposed to reduce the complexity and peak-to-average power ratio (PAPR), and improve carrier frequency offset (CFO) robustness for OFDMA systems. However, Q-OFDMA receiver obtains frequency diversity at the cost of noise enhancement. This paper proposes an iterative (turbo) equalization in conjunction with channel estimation for Q-OFDMA systems to mitigate the noise enhancement effect and improve the BER performance. In the proposed scheme, the channel estimation technique makes use of both training symbols and soft coded data information to suppress the inter-symbol interference (ISI) caused by channel estimation errors in Q-OFDMA systems. Simulation results show that performance improvement can be achieved with the proposed algorithms.

Extrinsic information transfer (EXIT) charts are utilized to optimize the iterative multiuser detector receiver in a multiuser turbo-coded CDMA system. The (receive) power levels are optimized for the system load using a constrained nonlinear optimization approach. The optimal decoding schedule is derived dynamically using the power optimized EXIT chart and a Viterbi search algorithm. Dynamic scheduling is shown to be a more flexible approach which results in a more stable QoS for a typical system configuration than one-shot scheduling, and large complexity savings over a receiver without scheduling. We verify through simulations that complexity savings of over 50% and power savings of over 8dB can be achieved. We show that the optimized power levels combined with adaptive scheduling allows for efficient utilization of receiver resources for heavily loaded systems.

In this letter, we propose a new multi-stage LLR combining (MLC) algorithm in an iterative receiver for MIMOBICM systems. This algorithm combines the soft information from different receive antennas in a multi-stage fashion, where the combining factors are derived based on the joint likelihood function of bivariate Gaussian random variables. The variance transfer (VT) function of the proposed scheme is derived for performance analysis. For slow fading channels, we show that the proposed MLC algorithm can achieve almost the same performance as the linear minimum mean square error (LMMSE) filtering approach, whereas the computation-demanding matrix inverse for LMMSE can be avoided.

Coherent OFDM detection requires accurate channel state information (CSI). Mobile radio channels are both time and frequency dispersive, especially at high vehicular speeds, which makes channel estimation a challenging problem in system design. Conventional preamble-based and pilot-aided channel estimation require numerous reference signals, which significantly compromises the system throughput. This paper proposes a novel low complexity iterative turbo channel estimation technique which makes use of preamble, pilots and soft decoded data information in an iterative fashion to improve the system performance over the time and frequency selective fading channel while maintaining the system throughput. The numerical and analytical results show that the proposed technique can approach the performance of systems with perfect CSI with much fewer preamble and pilots symbols compared to existing channel estimation methods.

A joint iterative channel estimation and data detection receiver is proposed in this paper for multiple-input- multiple-output orthogonal frequency deivision multiplexing (MIMO-OFDM) system, which employs two mandatory MIMO profiles as in IEEE 802.16e MIMO mobile WiMAX system, namely Alamouti space time coding (STC) and spatial multiplexing (SM), over time and frequency dispersive fading channel. A novel channel estimator with time-domain and frequency-domain combining of channel estimates from preamble/pilot/soft decoded data information is proposed and integrated with maximum ratio combining (MRC) data detector and co-antenna interference canceler (IC) for STC system and SM system respectively. Numerical results show that the proposed receiver has 2dB gain compared to conventional non-iterative receivers in pedestrian low mobility condition and more than 6dB gain in vehicular high Doppler environment.

In this paper we derive a novel extrinsic information transfer (EXIT) function for a channel estimator that uses both pilot symbols and soft decoding information. We consider a CDMA system with forward error correction that employs iterative multiuser detection based on soft interference cancellation and single user decoding. The J function is useful in convergence analysis as it provides a mapping between mutual information and variance. J functions are derived for both the flat-fading Rayleigh channel and channel estimate, allowing the a priori inputs to the channel estimator, interference canceller and decoder to be modeled. Algorithms for obtaining the EXIT functions for the three receiver blocks are proposed. We then combine the EXIT functions to create an EXIT chart that can be used to predict the behaviour of the overall receiver. Simulations are run for a 3GPP-like receiver in a mobile environment with a vehicle speed of 100 km/h. It is found that under these conditions, the detector trajectory closely matches our proposed EXIT chart.

In this paper, we propose a new linear detector for the iterative receiver for MIMO-BICM systems. Instead of performing vector matched filtering after the interference cancellation, the proposed detector successively combines the soft information from different receive antennas, where the combining factors are computed based on the joint likelihood function of multivariate Gaussian random variables. Numerical results show that significant performance improvement can be achieved and the additional complexity is negligible.

In current OFDMA systems, three major problems arise due to the large number of subcarriers, including high peak to average power ratio (PAPR), sensitivity to carrier frequency offset (CFO), and high complexity in user terminal. In this paper, based on an innovative concept ofLayeredFFTstructure, we propose novel quadrature OFDMA system which can overcome these problems. In particular, the proposed system can achieve same guard-interval overhead and same bandwidth occupation to conventional OFDMA schemes, while with significantly reduced complexity and power consumption in user terminals. Simulation results show that the proposed quadrature OFDMA system can achieve similar and even better performance than conventional OFDMA systems, particularly when a minimum mean square error (MMSE) equalizer is applied.

In this paper, we develop low-complexity iterative channel estimation techniques for emerging IDMA systems. The channel estimators make use of pilots as well as soft decoded data information. We derive a lower bound for channel estimation error that reflects the reliability of soft decoded data. We show that the estimators perform close to a minimum variance unbiased estimator as the mean square error (MSE) approaches the lower bound. Numerical results on the MSE and BER performance also show that the proposed channel estimators are able to track the time-varying channel states.

Current OFDM systems assume the channel is not time varying within one OFDM frame, and use channel estimates obtained from preamble or pilots for data symbols within the same frame. This performs poorly in rapid dispersive fading channel with high mobility. This paper proposes a novel iterative turbo channel estimation to improve the performance of OFDM system over fading channel with both time and frequency selectivity. Unlike the conventional channel estimation techniques only using preamble or pilots, the proposed channel estimation technique makes use of preamble, pilots and soft coded data information in a turbo iterative approach. Mean square error analysis and simulation results show that the proposed channel estimation technique can approach the performance of perfect channel state information (CSI) with fewer pilots insertion.

In this paper the system model of an uncoded and fully synchronised interleave division multiple access (IDMA) scheme is transformed into an equivalent matrix model. This new model allows for additional alternative insights into the principles of IDMA and it is shown that the arising matrices are highly structured and very sparse. Furthermore the optimal detector for IDMA is derived and compared to the standard iterative decoder. The authors use exit charts and an empirical rank analysis to highlight that systems without an alternating signature mask always have a low rank and that parallel interference cancellation then fails.

This paper proposes a high-performance and relatively low-complexity approximated maximum likelihood estimator (MLE) for carrier frequency offset (CFO) in practical orthogonal frequency multiplexing (OFDM) systems, where the preamble has repetition structure in time domain, null subcarriers are allocated at two ends of the frequency spectrum, and the channel impulse response is assumably not longer than the cyclic prefix. We show the optimality of conventional MLEs do not hold in such systems, and derive a universal MLE and its modified Cramer-Rao lower bound in the general case. To reduce the complexity, we propose to cascade it with a low-complexity rough CFO estimator so that only a moderate number of maxima searches are needed to achieve satisfactory performance. Simulation with the IEEE 802.16 preamble symbol confirms the excellent performance of the proposed method.

In this paper, we investigate the performance of DS-CDMA systems in dispersive fading channels. We develop a low-complexity receiver that jointly performs channel estimation and multiuser decoding in an iterative manner. Pilot symbols, together with the soft decoding information on the transmitted data are harnessed to estimate the time-varying channel states, which in turn leads to improved detection and decoding results. We show through simulation and mean square error analysis that the proposed system can sustain a high-throughput WCDMA transmission.

In this paper, we propose a low-complexity multistage filter for the partitioned-spreading CDMA (PS-CDMA). The proposed method embodies a MMSE filter that operates on the soft output of interference cancellation. Extrinsic information on the partitioned symbol is exchanged between the MMSE filter and the partitioned-spreading APP detector to improve the system performance. It is shown via variance evolution the signal-to-noise ratio of this novel technique achieves that of the single-user channel even for heavily loaded systems. It also outperforms IDMA by a fraction of dB to a few dBs, for a variety of system loads. The complexity of the proposed system is much lower that of its counterparts for conventional CDMA, which makes it a viable alternative for future 3G or beyond 3G systems

Iterative interference cancellation of coded code-division multiple access (CDMA) using random spreading with linear cancellation is analyzed. If users are grouped into power classes and Shannon bound approaching codes are used, a geometric power distribution achieves the additive white Gaussian noise (AWGN) channel Shannon bound as the numbers of classes becomes large. The optimal distribution of the size of these classes is shown to be uniform. If users are grouped into different rate classes with equal powers among equal rate users, the Shannon bound for AWGN channels can be achieved with an arbitrary distribution of the classes sizes, provided that the size of the largest rate class obeys the mild condition that its ratio of size to processing gain is much smaller than the inverse of the signal-to-noise ratio (SNR). The case of equal powers and equal rates among all users is addressed as a "worst case" scenario. It is argued that simple repetition codes provide for a larger achievable capacity than stronger codes. It is shown that this capacity monotonically increases as the rate of the code decreases. A density evolution analysis is used to show that the achievable rates exceed those of a minimum-mean square error filter applied to the uncoded signals. This lower bound is tight for small ratios of bit energy to noise power, and otherwise the iterative cancellation receiver provides an appreciably larger capacity. Relating to recent result from the application of statistical mechanics it is shown that the repetition-coded system with iterative cancellation achieves the performance of an equivalent optimal joint detector for uncoded transmission

In this paper, we develop a new soft-in soft-out (SISO) multiple-input multiple-output (MIMO) detection algorithm using the Markov chain Monte Carlo (MCMC) simulation techniques and study its performance when applied to a MIMO communication system. Comparison with the best MIMO detection algorithm in the current literature, sphere decoding, shows that the proposed detection algorithm can improve the gap between the present results and the capacity by as much as 2 dB.

Multiple-access interference (MAI) has been considered as a major performance-limiting factor in the next-generation CDMA systems. Multiuser detection (MUD) methods have been proposed to mitigate the MAI from the co-channel users by incorporating the cross-correlation properties between users. Recently, two classes of emerging techniques, probabilistic data association (PDA) and Markov Chain Monte Carlo (MCMC) methods, have been applied to the multiuser detection. In this paper, we present a new method, named Monte Carlo PDA (MC-PDA), that incorporates the concepts of both to give a more reliable inference of the CDMA symbols by appropriately modelling and updating the MAI. The methodology is general and can be applied to other communication channels

We present a novel soft-in soft-out (SISO) detection scheme based on Markov-chain Monte-Carlo (MCMC) simulations. The proposed detector is applicable to both synchronous multiuser and multiple-input multiple-output (MIMO) communication systems. Unlike previous publications on the subject, we use Monte Carlo integration technique to arrive at the receiver structure. The proposed multiuser/MIMO detector is found to follow the Rao-Blackwell formulation and found to result in much more reliable systems than those reported so far in the literature. In particular, through computer simulations, we demonstrate that the proposed receiver can easily handle the cases of overloaded systems, i.e. when in a multiuser set-up the number of users is larger than the spreading gain, or when in a MIMO set-up there are more transmit antennas than receive antennas.

The combination of forward error control coding (FEC) with code division multiple access (CDMA) using random spreading sequences is considered. Through tracking the input-output variance evolutions of the decoding components, which are the CDMA interference resolution function and the individual error control decoders, an analysis of the iterative joint detection was previously presented. It was shown this variance transfer (VT) analysis between component decoding devices gives very accurate description of the convergence properties of the iterative joint detector for a CDMA system with all users received at equal energy. In this paper, the VT analysis is extended to the more realistic cases where the different users have unequal power levels. It is shown that the effective VT functions modified w.r.t. the unequal power scenario can be used to analyze the iterative reception employing a wide variety of linear filters for interference resolution, in a manner very similar to that in the equal power case. Numerical simulations verify the results predicted by VT analysis.

Low complexity multistage filters are inserted into the soft-information exchange paths of an iterative joint detector for error control coded CDMA signals. It is shown that these filters approach the performance of the optimal, but complex, per-user minimum-mean square error filter with a few stages. A variance transfer analysis is presented to quantify the performance of this method and to study the case of unequal received power distributions.

A code-division multiple-access (CDMA) system in conjunction with serial concatenation of FEC codes is considered. Iterative decoding strategies are used at the receiver. The component decoders are functionally described by variance transfer (VTR) charts. Pinch-off SNRs, where the iterative decoding starts to function, are precisely predicted by this approach. Designs of the multiuser detection (MUD) method and the component FEC codes have been carried out and system performance to within about 1 dB of the Shannon capacity of the multiple-access channel is demonstrated.

Joint iterative decoding of forward error control (FEC) encoded code-division multiple-access (CDMA) systems is studied. Serially concatenated convolutional codes are used as FEC codes. We apply iterative decoding to this system and study the effect of component decoders and different decoding partition schemes via a variance transfer analysis. It is shown that distinct partition schemes result in similar pinch off SNRs.

Joint iterative decoding of multiple forward error control (FEC) encoded data streams is studied for linear multiple access channels, such as code-division multiple access (CDMA). It is shown that such systems can be viewed as serially concatenated coding systems, and that iterative soft-decision decoding can be performed successfully To improve power efficiency, powerful FEC codes are used. These FEC codes are themselves serially concatenated. The overall transmission system can be viewed as the concatenation of two error control codes with the linear multiple access channel, and soft-decision decoders are used at each stage. A variance transfer function approach applied to the analysis of this system captures the role of the component decoders in an overall iterative decoding system. We show that this approach forms a methodology to study the effects of the component codes as well as that of the iteration schedule. Analysis and simulation examples are presented for transmission systems that operate close to the Shannon limit and illustrate the accuracy of the analysis.