27 February 2018 in IEEE Transactions on Communications
Relay transmissions play an important role in many types of communication systems. In this paper, we consider packet-level coded transmissions over lossy relay links with the finite-buffer and coding coefficients delivery cost constraints. We propose a low-complexity coding scheme where packets are encoded from sequentially formed random subsets of source packets called batches. The relay recodes only from the buffered packets belonging to the same batch to maintain the code sparsity, lowering the packet header overhead and the decoding complexity compared to the random linear network coding (RLNC). To analyze the performance, we first propose an absorbing Markov chain model to analyze the RLNC transmission over finite-buffer relay links. The finite-length analysis not only provides a lower bound on the completion time using any sparser random codes but also characterizes each individual batch's transmission of the proposed code. Based on the analysis, another Markov chain is proposed to determine the decoding failure probability and the expected completion time of the batch coding scheme. As shown through analysis and simulations, the proposed scheme achieves higher effective end-to-end rates than RLNC when the coding coefficients delivery cost is considered and is also with much lowered decoding complexity thanks to its sparseness.
1 January 2011 in Progress In Electromagnetics Research
In this paper, microstrip line resonators with loaded elements are proposed and studied to design microstrip diplexer. To demonstrate the design ideas, the equivalent circuits of the proposed resonators are built and studied. It is found that the different loads on different positions of the proposed half-wavelength resonator make the resonator have different features, which will easily control the characteristic of the diplexers. And here, resistor, open stub, and shorted stub are used as loaded elements. It is found the resistor loaded on the center of the microstrip line resonator can extremely reduce the unloaded quality factor of even-mode resonant frequency, which can be used to suppress the harmonics of the diplexer. The loaded open stub not only can reduce the size of the diplexer, but also can control the frequency ratio between the fundamental frequency and second harmonic of a resonator, which can increases the frequency ratio between the two passbands of the diplexer. As for the loaded shorted stub, it can enlarge the size of the diplexer. To demonstrate the design ideas, three diplexers are presented. The comparisons between the loaded and unloaded diplexers are given. The experimental results agree well to the theoretical predictions and simulations.
29 June 2020 in IEEE Transactions on Vehicular Technology
This paper compares the end-to-end (E2E) fountain codes and random linear network coding (RLNC) for lossy multi-hop short-packet communications, under the constraint that the coding coefficient overhead imposed by RLNC recoding is non-negligible. The decoding failure probability (DFP) of RLNC is analyzed and computed as a function of the most relevant parameters. A DFP minimization problem is formulated and numerically solved from a cross-layer perspective. By comparing the codes using optimized parameters, the DFP of RLNC is shown to be higher than that of the E2E codes, which is due to the coefficient overhead. The results suggest that E2E fountain codes might be preferred for multi-hop short-packet communications.
13 March 2020 in IEEE Transactions on Wireless Communications
Data dissemination to multiple users within a predetermined deadline is a commonly required function in many emerging wireless scenarios. In this paper, we propose a data dissemination scheme enhanced by device-to-device (D2D) communications and random linear network coding (RLNC), where the broadcast and the D2D links are used simultaneously with on-the-fly RLNC to accelerate the dissemination. The probability distribution of the completion time of using either pure or systematic RLNC is analyzed assuming a finite field size, under a general model where the cooperation may occur probabilistically and multiple packets may be exchanged between the D2D users in each time slot. An optimization problem is formulated to design the cooperation parameters to minimize the expected energy consumption of the system. Numerical and simulation results show that the analysis is accurate, and that to minimize the energy consumption a tradeoff exists between allowing for longer broadcast and for more intensive cooperation. The cooperation parameters designed for multiple D2D pairs are shown to be effective for meeting the reliability requirements.
24 November 2016 in IEEE Communications Letters
This letter proposes a sparse network code based on band matrices for lossy networks. The proposed code is precoded, has no wrap-around, and allows only a limited number of bands. Compared to existing works, the code simplifies decoding, enables more flexible recoding, and is able to significantly reduce the reception overhead that might be caused by recoding.
12 December 2022 in IEEE Canadian Journal of Electrical and Computer Engineering
Radio frequency fingerprint (RFF) based authentication of wireless devices can be used in the fields of the access security at the physical-layer of wireless networks and radio spectrum management. However, the stability of RFF is easily damaged by the wireless multipath fading channel in mobile communications. An RFF fingerprinting method with the nonlinearity and in-phase and quadrature (IQ) imbalance of the transmitter is proposed for chunk-based wireless orthogonal frequency division multiplexing (OFDM) devices based on a Hammerstein system parameter separation technique, which cancels the adverse influence of the time-varying multipath channel. First, the parameters of the nonlinear model of the transmitter and finite impulse response (FIR) of the wireless multipath channel are estimated with the Hammerstein system parameter separation technique. Second, the best IQ imbalance parameter combination is obtained with the FIR estimation of the channel. Finally, the estimated parameters of the nonlinear model and IQ imbalance are used as RFFs to classify the transmitters. Theoretical analyses and numerical experiments demonstrate that the obtained RFFs are stable and the fusion authentication of transmitters with subtle differences from the same model and same series is feasible using the novel method.