Jiaru Lin

A joint recovery algorithm for distributed compressed sensing

Distributed compressed sensing exploits the correlation among multiple signals to reduce the number of measurements required for recovery. In this paper, we propose a recovery algorithm for a type of joint sparsity model, where all signals share a common sparse component and each individual signal contains a sparse innovation component. Our approach iteratively removes the information of each component from the measurements and then performs sparse recovery. We provide analytical analysis to verify the advantage of the proposed algorithm over separate recovery, which is also confirmed by simulation results. Copyright

A channel estimation method based on distributed compressed sensing and time-domain Kalman filtering in OFDM systems

Channel estimation is important for coherent detection in orthogonal frequency-division multiplexing (OFDM) systems. Current time-domain Kalman filtering (TDKF) method has a good performance in estimating the channel responses, but is impractical since it requires the knowledge of multipath delays. In this paper, we propose a new scheme to relax such requirement by combining the recent methodology of distributed compressed sensing (DCS) and TDKF. By exploiting the sparse attribute of OFDM channels, the number of pilots could be reduced greatly. Furthermore, to reduce the complexity, a threshold on the change of channel responses is designed to avoid unnecessary DCS execution. Simulations indicate the proposed method achieves better performance than conventional least square method.

Connectivity of interference limited cognitive radio networks

This paper considers the connectivity of cognitive radio networks, which is defined as the existence of an infinite connected component of secondary users. The connectivity reveals the possibility of communication between an arbitrary pair of nodes in the secondary network, while long range communication is achieved in a multi-hop fashion. In their seminal work, Wei Ren et al. introduce the concept of connectivity region, and characterize some basic properties based on the Poisson Boolean model. In this paper, we take interference into consideration and study the impact of interference to the connectivity of the secondary network. Using methodologies of stochastic geometry and percolation theory, we find that the connectivity region without taking interference into consideration is a subset of our connectivity region, which means interference will not adversely affect the connectivity of secondary network. Specifically, as long as the density pair of the two networks lies in the connectivity region established in the Poisson boolean model, there must exist an orthogonal factor γ which guarantees the connectivity of the interference limited secondary network.

Sub-Sampling Framework of Distributed Video Coding

Distributed video coding (DVC) has recently been proposed to reduce the complexity of the encoder, whereas it suffers from the sampling cost of huge amount of image data. To relax such sampling burden, this paper develops a novel sub-sampling distributed video coding (SuDVC) by utilizing compressive sensing (CS) technique. Due to the inherent sparsity in video sources, the video frames are compressively sampled at the encoder. On the other hand, by exploiting the correlation between CS measurements and side information and by performing sparsity recovery, the video frames are recovered at the decoder. When compared with the traditional fully-sampling equivalence, SuDVC enjoys the reduction of transmission rate, the reduction of implementation complexity and the robustness to channel losses, which are verified in the simulations.

Transmission capacity of secondary networks in hybrid overlaid/underlaid cognitive radio systems

We study the transmission capacity of the secondary (SR) network in the hybrid cognitive radio (CR) system where underlaid and overlaid access approaches are combined to improve the transmission capacity of the SR network. The transmission behavior of the primary (PR) network is characterized as a two-state (idle and busy) discrete time Markov model, and sensing performance of the SR network captured by the missed detection probability and false alarm probability is also considered. We analyze their effects on the transmission capacity of the SR network and the outage probability of the PR network. Analysis and numeric results show that hybrid mode has greater superiority than either overlaid mode or underlaid mode on the transmission capacity of the SR network, especially when the PR network has light load. The missed detection has good effect on the transmission capacity of the SR network, but increases the outage probability of the PR network, which is dispointed. The false alarm has no effect on the outage probability of the PR network, but decreases the transmission capacity of the SR network.

Rate control for network coding based multicast: a hierarchical decomposition approach

In this work we consider the rate control issue for network coding based multicast among multiple sessions, which can be formulated as a network utility maximization problem. To solve the problem we propose a distributed optimization decomposition approach, which is different from the previous work in the literature in that (1) it is a hierarchical decomposition approach where the primal problem is decomposed recursively, until an independent rate control module is obtained and the decomposed subproblems can be solved by the distributed max-flow algorithm and we emphasize the layered functionality allocation of decomposed subproblems following the framework of Layering as Optimization Decompositions; (2) we first separate the primal problem by relaxing the capacity constraint among sessions; (3) to implement end-to-end control, we separate the independent rate control module at end node from the operation in the interior of the network. In this work we intend to propose not only a rate control algorithm but also a possible choice of network architecture for network coding based communication with rate control capability.

Algorithms for optimal resource allocation in heterogeneous cognitive radio networks

In this paper, the problem of resource allocation in heterogeneous cognitive radio networks is investigated under the constraints of interference temperature limit. The resource allocation problem is formulated as a mixed-integer programming problem and solved by Lagrangian dual method based on which a centralized subgradient update algorithm is proposed. The approximate optimality of this algorithm is promised by time-sharing condition. We also realize this centralized algorithm distributively following the implication of dual decomposition. Two algorithms are compared in terms of complexity, communication signaling, latency etc.. At last, simulations are carried out to validate the proposed algorithms.

Relay beamforming with power minimization in cognitive radio network

This paper studies the cognitive one-way relay network which is composed of a pair of primary transmitter and receiver. A two-step one-way amplify-and-forward (AF) relaying scheme is used by the primary network who utilizes the cognitive radio (CR) terminals as its relay nodes. Simultaneously, the cognitive information is transmitted by multiple CR terminals to the CR base station. In the first phase of AF relaying, the primary signal is forwarded by transmitter to the CRs/relays. The CRs/relays multiple the cognitive signals to the received primary signal and accomplish a distributed beamforming in the second phase of relaying. The authors aim at minimizing the CRs/relays power under the different signal to interference pulse noise ratio (SINR) restraints at the primary receiver and the cognitive base station (BS). The results can be seen from the simulation that the minimum CR terminals/relays power will increase with the rising of the SINR constraints. Besides, the performance difference between CR terminals/relays beamforming and without beam-forming is affected by the minimum CR terminals transmit power and the number of relays. Furthermore, with the number of CR terminals/relays rising, the performance difference is decreasing.

User scheduling schemes based on limited feedback in cognitive radio networking

In this paper, a joint user scheduling and beamforming method based on limited feedback is studied for a cognitive radio (CR) multi-user multi-input multi-output (MU-MIMO) network, where a large number of secondary users (SUs) coexist with a primary user (PU). In the proposed approach, secondary users feed precoding vector indices (PVIs) and signal to interference plus noise ratio (SINR) related channel quality information (CQIs) back to the secondary base station (SBS), and then SBS chooses appropriate SUs according to the knowledge about the link information from itself to the primary user and the quality of service (QoS) requirement of the PU. Simulation results show that the proposed algorithm can achieve good performance in a robust manner.

A zero-forcing method for downlink in cognitive multiuser MIMO system

One precoding algorithm used in the downlink of CR MU-MIMO system is presented. Though interference between CUs is not mitigated completely, co-channel interference to PU is pre-whitened and various performance evaluation results demonstrate that algorithm CR-TR-BD accommodates for more general case and is able to achieve high sum rate throughput than BD only used in CR network.