An Unequal Error Protection Scheme using Parallel Concatenated Low-Density Parity-Check Code

Transmitting visible applications over wireless networks is a challenging work. Bandwidth, channel noise, and heterogeneous receivers are critical factors which will affect the quality of visible applications. Thus cross-layer design of wireless network is proposed to address these problems. In general cross-layer design, Unequal Error Protection(UEP) is an important work to address noisy transmission channel and adaptive transmission rate. So we want to propose an UEP scheme using Parallel Concatenated LDPC Code(PCGC). PCGC is a powerful code whose BER performance is better than LDPC code under some constraints. In this thesis, we propose a new kind of stopping criterion to reduce long latency of PCGC decoding. Besides, we introduce some puncturing methods for finite length LDPC code and propose a puncturing scheme for PCGC system. Finally, we will propose an UEP scheme which uses PCGC as its error-correcting code. In the work of stopping criteria, we want to propose a local stopping criterion to couple with existing global stopping criteria. In the work of puncturing, we let the mentioned puncturing methods to select recommended punctured-bits in individual constituent codes of PCGC, and then to propose how to select punctured-bits over all PCGC system. In the final work, we will propose an application and channel aware UEP scheme in the viewpoint of relation between frame PSNR and decoded BER. Finally, we will show the performances to verify advantages of our works.1 Introduction 1 1.1 Why Joint Source-Channel Coding? . . . . . . . . . . . . . . . . . . . 1 1.2 Cross-Layer Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 Related Works 4 2.1 Low-Density Parity-Check Code . . . . . . . . . . . . . . . . . . . . . 5 2.1.1 Sparse Parity-CheckMatrix and Tanner Graph . . . . . . . . 5 2.1.2 Massage-Passing Algorithm . . . . . . . . . . . . . . . . . . . 7 2.2 Parallel Concatenated Gallager Code . . . . . . . . . . . . . . . . . . 9 2.2.1 PCGC Encoder . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.2.2 PCGC Decoder . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.3 Existing Global Stopping Criteria . . . . . . . . . . . . . . . . . . . . 15 2.3.1 Cross-Entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.3.2 Sign-Change-Ratio . . . . . . . . . . . . . . . . . . . . . . . . 18 2.3.3 Hard-Decision-Aided . . . . . . . . . . . . . . . . . . . . . . . 19 2.3.4 Simulations and Summary . . . . . . . . . . . . . . . . . . . . 20 3 Proposed Stopping Criteria 23 3.1 Proposed Local Stopping Criteria . . . . . . . . . . . . . . . . . . . . 24 3.1.1 Agreement of Reliability-Trend . . . . . . . . . . . . . . . . . 25 3.1.2 Variation of Reliability-Trend . . . . . . . . . . . . . . . . . . 28 3.1.3 Local HDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.2 Combine Global and Local Stopping Criteria . . . . . . . . . . . . . . 35 3.2.1 Feasibility of using Cross-Entropy . . . . . . . . . . . . . . . . 36 3.2.2 Feasibility of using Sign-Change-Ratio . . . . . . . . . . . . . 39 3.2.3 Feasibility of using Hard-Decision-Aided . . . . . . . . . . . . 42 3.3 Simulations and Summary . . . . . . . . . . . . . . . . . . . . . . . . 44 4 Unequal Error Protection using PCGC 50 4.1 Puncturing for Finite Length LDPC Codes . . . . . . . . . . . . . . . 51 4.1.1 Step-Recoverable . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.1.2 Node Distance Constraint . . . . . . . . . . . . . . . . . . . . 54 4.2 Puncturing of PCGC . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 4.2.1 Puncturing at lower-MCWpart . . . . . . . . . . . . . . . . . 56 4.2.2 Puncturing at higher-MCWpart . . . . . . . . . . . . . . . . 58 4.2.3 Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 4.3 Proposed UEP using PCGC . . . . . . . . . . . . . . . . . . . . . . . 64 5 Simulation and Analysis 70 5.1 PCGC Performance Analyses . . . . . . . . . . . . . . . . . . . . . . 70 5.2 Puncturing Performance . . . . . . . . . . . . . . . . . . . . . . . . . 74 5.3 UEP using PCGC: A Case Study on H.264 . . . . . . . . . . . . . . . 74 6 Conclusion and Future Work 8