Enhanced Codebook of Sparse Vector Coding Based on Mean-Variance Trade-Off Model for URLLC

• Published in: IEEE communications letters Vol. 29; no. 6; pp. 1310 - 1314
• Main Authors: Yang, Yifei, Chen, Changju, Zhu, Pengcheng
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.06.2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; codebook; Coding; Compressed sensing; Computational complexity; Convexity; Correlation coefficient; Correlation coefficients; Decoding; Dictionaries; Encoding; encoding optimization; Integer programming; Mean; mean-variance trade-off; MPT; Optimization; Packet transmission; SDP; Semidefinite programming; Simulation; Sparse vector coding; Static VAr compensators; Tradeoffs; Ultra reliable low latency communication; URLLC; Variance; Vectors
• ISSN: 1089-7798; 1558-2558
• DOI: 10.1109/LCOMM.2025.3559985

Sparse Vector Coding (SVC) is a novel coding scheme of short packet transmission in Ultra-Reliable Low-Latency Communication (URLLC). SVC is usually modeled as a standard Compressed Sensing (CS) model, so the column correlation coefficient of the encoding dictionary will directly determine the decoding performance. Starting from the point of view of optimizing codebook, this letter will first model the minimization of the maximum column correlation coefficient as a linear integer programming (LIP) problem, and obtain a tighter solution than existing studies. Then, the optimization objective was transformed into statistical parameters of column correlation coefficient distribution and modeled as mean-variance trade-off model, which was a convex optimization problem and optimized by Semi-Definite Programming (SDP) and Modern Portfolio Theory (MPT) respectively, improving the Block Error Ratio (BLER) performance about 1dB and reduced the computational complexity. Simulation results verify the effectiveness of the above algorithms and improve the decoding performance effectively.