Theme detection is a fundamental task in user-centric dialogue systems, aiming to identify the latent topic of each utterance without relying on predefined schemas. Unlike intent induction, which operates within fixed label spaces, theme detection requires cross-dialogue consistency and alignment with personalized user preferences, posing significant challenges. Existing methods often struggle with sparse, short utterances for accurate topic representation and fail to capture user-level thematic preferences across dialogues. To address these challenges, we propose CATCH (Controllable Theme Detection with Contextualized Clustering and Hierarchical Generation), a unified framework that integrates three core components: (1) context-aware topic representation, which enriches utterance-level semantics using surrounding topic segments; (2) preference-guided topic clustering, which jointly models semantic proximity and personalized feedback to align themes across dialogue; and (3) a hierarchical theme generation mechanism designed to suppress noise and produce robust, coherent topic labels. Experiments on a multi-domain customer dialogue benchmark (DSTC-12) demonstrate the effectiveness of CATCH with 8B LLM in both theme clustering and topic generation quality.

User simulators are crucial for replicating human interactions with dialogue systems, supporting both collaborative training and automatic evaluation, especially for large language models (LLMs). However, current role-playing methods face challenges such as a lack of utterance-level authenticity and user-level diversity, often hindered by role confusion and dependence on predefined profiles of well-known figures. In contrast, direct simulation focuses solely on text, neglecting implicit user traits like personality and conversation-level consistency. To address these issues, we introduce the User Simulator with Implicit Profiles (USP), a framework that infers implicit user profiles from human-machine interactions to simulate personalized and realistic dialogues. We first develop an LLM-driven extractor with a comprehensive profile schema, then refine the simulation using conditional supervised fine-tuning and reinforcement learning with cycle consistency, optimizing at both the utterance and conversation levels. Finally, a diverse profile sampler captures the distribution of real-world user profiles. Experimental results show that USP outperforms strong baselines in terms of authenticity and diversity while maintaining comparable consistency. Additionally, using USP to evaluate LLM on dynamic multi-turn aligns well with mainstream benchmarks, demonstrating its effectiveness in real-world applications.

In this paper, the physical layer security is studied in a buffer-aided MIMO relaying system, which consists of a source, a destination, a multi-antenna relay and multiple eavesdroppers. The relay operates in either half-duplex mode or full-duplex mode, based on the channel state information (CSI) of legitimate links. The exact closed-form expressions for the secrecy throughput and the outage probability of the buffer-aided MIMO relaying system are derived. In this work, the transmission rates can be adjusted at the source and relay instead of the predetermined transmission rates in the existing work. The optimal transmission rates are obtained to maximize the average secrecy throughput under the intended secrecy outage probability constraint, where only statistical CSI is needed when the optimization problem is solved. Numerical results show that analytical results of the secrecy outage probability agree with simulation results well. Compared with the conventional bufferless relaying schemes, the proposed buffer-aided relaying scheme achieves a higher throughput, as well as preserving the secrecy performance.

In this study, the authors analyse the ergodic capacity of the downlink channel in a hybrid satellite–terrestrial cooperative system. The satellite link experiences shadowed Rician fading and the terrestrial link follows Rayleigh fading. Using the moment generating function‐based approach, they derived the formulae for the ergodic capacity with the amplify‐and‐forward protocol or the decode‐and‐forward protocol. Moreover, the asymptotic expression for the ergodic capacity at a high signal‐to‐noise ratio is derived, and the system multiplexing gain is also obtained. Using the asymptotic results, they propose a simple relay selection strategy, which only requires the knowledge of channel statistics. Monte Carlo simulations are provided to confirm the analytical results, and to illustrate the effect of different factors, in terms of the number of relays, channel conditions, and the cooperative protocol, on the ergodic capacity. Numerical results show that the analytical results are highly accurate in the single relay scenario and have low computational complexity in the multi‐relay scenario. In addition, the proposed relay selection strategy achieves low overhead in the price of a little performance degradation.

Cognitive radio is an effective technology to alleviate the spectrum resource scarcity problem by opportunistically allocating the spare spectrum to unauthorized users. However, a serious denial-of-service (DoS) attack, named the ‘primary user emulation attack (PUEA)’, exists in the network to deteriorate the system performance. In this paper, we propose a PUEA detection method that exploits the radio channel information to detect the PUEA in the cognitive radio network. In the proposed method, the uniqueness of the channel impulse response (CIR) between the secondary user (SU) and the signal source is used to determine whether the received signal is transmitted by the primary user (PU) or the primary user emulator (PUE). The closed-form expressions for the false-alarm probability and the detection probability of the proposed PUEA detection method are derived. In addition, a modified subspace-based blind channel estimation method is presented to estimate the CIR, in order for the proposed PUEA detection method to work in the scenario where the SU has no prior knowledge about the structure and content of the PU signal. Numerical results show that the proposed PUEA detection method performs well although the difference in channel characteristics between the PU and PUE is small.

A complete description of trust relationship is key to construct a high precision trust model. But most of existing models not only miss the negative information and the hesitation information of trust, but also ignore the discordance between text comments and ratings. To solve the problems, a dynamic trust evaluation model based on the affect intensity is proposed. In the model, the intensity of sentimental polarities are calculated from words in comments. The corresponding relation between evaluations of trust property and the emotional intensity vector can be also described. Time series weights to distinguish the importance of transaction at different time are determined by the inverse form of exponential distribution. So the dynamic attenuation of trust can be described. To improve the polymerization ability of trust information, the local trust, the feedback trust and the overall trust are calculated by operators of fuzzy logic. The experimental results show the proposed model can effectively describe the trust relationship between nodes to identify and eliminate various malicious attacks significantly. Copyright © 2016 John Wiley & Sons, Ltd.

In this paper, the ergodic channel capacity is analyzed for a hybrid satellite-terrestrial single frequency network (SFN), which consists of a satellite (the source node), a mobile terminal (the destination node), and several gap fillers (the relay nodes) on the ground. The satellite-relay (S-R) link and the satellite-destination (S-D) link experience independent Rician fading, and the relay-destination (R-D) link experiences Nakagami-m fading. Based on the classical conclusions in multipath fading channels, the exact closed-form expression of the ergodic channel capacity for a hybrid satellite-terrestrial SFN with the decode-and-forward cooperative protocol is derived. The numerical results agree well with the simulation results. As the total transmit power of all nodes is assumed to be limited, the performance of hybrid satellite-terrestrial SFN is better than that of the no-relay system when the S-D link is in heavy-fading.

Using differential transmission method, the space-time block codes (STBC) can be applied in a complicated communication environment without channel state information (CSI). However, since the differential calculation includes multiplications, additions and conjugations, differential space-time block codes (DSTBC) also cause constellation expansion and result in nonlinearity problems in a wireless transmitter. In this paper, we proposed a new double mapping method for the DSTBC. In our proposed scheme, the attribute of pre-transmitted data bits is analyzed, and one of two mappings is chosen to generate current transmitted signal block. The proposed scheme also takes the advantage of differential encoding, and then CSI is not needed at the transmit antennas or at the receive antennas. Furthermore, the proposed DSTBC scheme can achieve full code rate without constellation expansion. Simulation results show that the performance of our proposed DSTBC scheme is close to that of traditional DSTBC scheme at a high SNR environment. Among all of the exited codes without constellation expansion, our proposed DSTBC has the minimum mapping table size and the best performance.

The coexistence and integration of multiple radio access technologies (RATs) deployed by different network operators are fundamental for next-generation mobile communication systems. In heterogeneous wireless environments, the access network selection, which is to select the most suitable network(s) for a terminal equipped with multiple radio interfaces according to various selection strategies, is a very important issue resulting in the performance maximization. In this paper, we propose a terminal-controlled access network selection scheme with parallel transmission capability for heterogeneous wireless networks, in which the suitable network(s) is(are) selected to access according to the total utility of different access networks. And the utility of each RAT is calculated based on the user preference and multiple access network criteria. Furthermore, the terminal-controlled access network selection with parallel transmission is formulated as a multi-constraint Knapsack problem which is solved with the exhaustive method to an optimum combination of RATs. Our proposed scheme is user-preference-aware, cost-aware, power-aware, and QoS-oriented. Simulation results show that our proposed access network selection scheme with parallel transmission can greatly improve the achieved data transmission rate with only a little increment in the cost and power consumption.

The wireless multiuser orthogonal frequency division multiple access (OFDMA) network in the fading channels with statistical quality of service (QoS) requirements is studied in this paper. Considering the statistical QoS requirement as a system performance metric, the concept of the effective capacity is introduced and the effective capacity region is characterized for a wireless multiuser OFDMA network. And then, we prove that the effective capacity region is convex. Moreover, the asymptotically optimal subcarrier and power allocation scheme which achieves the boundary points of the effective capacity region is proposed. Simulation results about the effective capacity region with two and three mobile stations under homogeneous and heterogeneous QoS constraints are given.

Linear combination-based collaborative spectrum sensing, in which a coefficient is used to weight each cognitive radio (CR) user's sensing result in the linear fusion process at the central controller, is an effective mechanism to solve the fading and hidden terminal problem in cognitive radio networks (CRNs). For a reliable scheme, it is important to distinguish whether the CR user is reliable or not. In this paper, we proposed a collaborative spectrum sensing scheme based on the ROCQ reputation management model for CRNs. In the proposed scheme, each CR user has a reputation degree used to calculate its coefficient in the linear fusion process, and the reputation degree is initialized and adjusted by the central controller according to each CR user's sensing result, sensing correctness and report consistency. Simulation results show that our proposed scheme alleviates the corrupted sensing problem resulting from the destructive channel conditions between the primary transmitter and CR users. Moreover, compared to the optimal linear fusion scheme, the detection performance of our proposed scheme is approximately same while it requires no instantaneous SNRs. And the convergence of the reputation degree adjusting method also performs well.

ABSTRACT The self‐healing group key distribution scheme with revocation can deal with the problem of distributing session keys for secure group communication over an unreliable wireless network, with the capability to resist packet loss and collusion attack. However, existing access‐polynomial‐based self‐healing key management schemes have some problems, such as the small number of active group members, much redundancy in the key updating broadcast packet, and limited collusion attack resistance capability. In order to increase the number of active group members and improve the performance of self‐healing group key distribution schemes, we propose a self‐healing group key distribution scheme based on the access polynomial and one‐way hash key chain for resource‐constrained wireless networks in this paper. In our proposed scheme, by binding the time at which the user joins the group with the capability of recovering previous group session keys, some novel methods to construct the personal secret, the access polynomial, and the session key updating broadcast packet are presented. Compared with existing schemes under same conditions, analysis and simulation results show that our proposed scheme not only supports much more group members and sessions, but also provides a stronger security, considering that it can deal with more colluding users. Moreover, our proposed scheme is especially suitable to resource‐constrained wireless networks in a very bad environment. Copyright © 2012 John Wiley & Sons, Ltd.

To provide the seamless mobility in heterogeneous wireless networks, the handoff decision strategy is regarded as a significant and challenging issue. In this paper, we propose a handoff decision algorithm with the parallel transmission characteristic of the mobile multimode terminal in heterogeneous wireless networks. In our algorithm, the model of Markov decision process is used to determine the optimal handoff policy in order to maximize the expected total reward during the transmission. Simulation results show that compared to some existing handoff decision algorithms, the proposed algorithm outperforms in terms of the expected total reward, the expected number of handoffs and the probability of connection dropping.

Due to the characteristics of the underwater acoustic channel, such as long propagation delay and low available bandwidth, the media access control (MAC) protocol designed for underwater acoustic sensor networks (UWASNs) is quite different from that for terrestrial wireless sensor networks. In this paper, we propose a MAC protocol for the UWASNs, named the funneling MAC (FMAC-U), which is a contention-based MAC protocol with a three-way handshake. The FMAC-U protocol uses an improved three-way handshake mechanism and code division multiple access (CDMA) based technology for request-to-send (RTS) signals transmitting to the sink in order that the sink can receive packets from multiple neighbors in a fixed order during each round of handshakes. The mechanism reduces the packet collisions and alleviates the funneling effect, especially alleviating the choke point of the UWASNs. Simulation results show that the proposed FMAC-U protocol achieves higher throughput, smaller packet drop ratio, lower end-to-end delay, and lower overhead of the control packet compared to the existing MAC protocols for UWASNs.

The 3780-point FFT is a main component of the time domain synchronous OFDM (TDS-OFDM) system and the key technology in the Chinese Digital Multimedia/TV Broadcasting-Terrestrial (DMB-T) national standard. Since 3780 is not a power of 2, the classical radix-2 or radix-4 FFT algorithm cannot be applied directly. Hence, the Winograd Fourier transform algorithm (WFTA) and the Good-Thomas prime factor algorithm (PFA) are used to implement the 3780-point FFT processor. However, the structure based on WFTA and PFA has a large computational complexity and requires many DSPs in hardware implementation. In this paper, a novel 3780-point FFT processor scheme is proposed, in which a 60×63 iterative WFTA architecture with different mapping methods is imported to replace the PFA architecture, and an optimized CoOrdinate Rotation DIgital Computer (CORDIC) module is used for the twiddle factor multiplications. Compared to the traditional scheme, our proposed 3780-point FFT processor scheme reduces the number of multiplications by 45% at the cost of 1% increase in the number of additions. All DSPs are replaced by the optimized CORDIC module and ROM. Simulation results show that the proposed 3780-point FFT processing scheme satisfies the requirement of the DMB-T standard, and is an efficient architecture for the TDS-OFDM system.

To provide reliable video stream transmission over wireless networks, we propose a cross-layer design model in this paper. In the proposed model, the adaptive modulation and coding parameters at the physical layer, the ARQ parameters at the data link layer and the application-layer forward error correction (A-FEC) parameters at the application layer are jointly designed. Our objective is to maximize the number of correctly decodable frames in a group of picture (GOP) according to the channel conditions and the GOP structure of the video stream. We also present a low-complexity bandwidth allocation algorithm for the A-FEC. The effectiveness of the proposed bandwidth allocation algorithm and the cross-layer design model was shown in the simulation results.

In this paper, retransmission strategies of the network-coding-based packet network are investigated. We propose two retransmission strategies, the packet-loss-edge-based retransmission strategy (PLERT) and the minimum retransmission strategy (MRT), which focus on optimizing the retransmission efficiency without the constraint on the encoding field size. We compared the performances of the proposed retransmission strategies with the traditional automatic repeat-request (ARQ) strategy and the random retransmission strategy. Simulation results showed that the PLERT strategy works well when the packet loss rate is small. Among these retransmission strategies, the performance of the MRT strategy is the best at the cost of the high complexity that is still polynomial. Furthermore, neither of the proposed strategies is sensitive to the encoding field size.