Scribd Logo
Upload

Welcome to Scribd!

  • 36 views

    Project: Ieee P802.15 Working Group For Wireless Personal Area Networks (Wpans)

    Uploaded by

    951m753m
    The document analyzes the performance and complexity of LDPC codes compared to convolutional and RS codes being considered for forward error correction in wireless networks. It presents simulation results showing that LDPC codes outperform convolutional codes in packet error rate across various code lengths and rates under an AWGN channel. It also finds that the coding gain difference between LDPC and convolutional codes increases with larger block sizes. Additionally, it compares the computational complexity of LDPC and RS decoding algorithms.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd

    Project: Ieee P802.15 Working Group For Wireless Personal Area Networks (Wpans)

    Uploaded by

    951m753m
    36 views12 pages
    The document analyzes the performance and complexity of LDPC codes compared to convolutional and RS codes being considered for forward error correction in wireless networks. It presents simulation results showing that LDPC codes outperform convolutional codes in packet error rate across various code lengths and rates under an AWGN channel. It also finds that the coding gain difference between LDPC and convolutional codes increases with larger block sizes. Additionally, it compares the computational complexity of LDPC and RS decoding algorithms.

    Original Title

    15 10 0241-00-004g Ldpc Code Performance and Complexity

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PPT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document analyzes the performance and complexity of LDPC codes compared to convolutional and RS codes being considered for forward error correction in wireless networks. It presents simulation results showing that LDPC codes outperform convolutional codes in packet error rate across various code lengths and rates under an AWGN channel. It also finds that the coding gain difference between LDPC and convolutional codes increases with larger block sizes. Additionally, it compares the computational complexity of LDPC and RS decoding algorithms.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    36 views12 pages

    Project: Ieee P802.15 Working Group For Wireless Personal Area Networks (Wpans)

    Uploaded by

    951m753m
    The document analyzes the performance and complexity of LDPC codes compared to convolutional and RS codes being considered for forward error correction in wireless networks. It presents simulation results showing that LDPC codes outperform convolutional codes in packet error rate across various code lengths and rates under an AWGN channel. It also finds that the coding gain difference between LDPC and convolutional codes increases with larger block sizes. Additionally, it compares the computational complexity of LDPC and RS decoding algorithms.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    of 12

    Sept 2009 doc.: IEEE 802.

    15-10-0241-00-004g
    Project: IEEE P802.15 Working Group for Wireless Personal Area Networks
    (WPANs)

    Submission Title: LDPC Code Performance and Complexity Comparing with Convolutional and
    RS Code
    Date Submitted: [01 Sept 2009]
    Source: [Qin Wang ] Company [University of Science & Technology Beijing]
    Address: [30 Xueyuan Road, Beijing 100083, China]
    Voice:[[+8610-62334781], FAX: [], E-Mail:[[email protected]]
    Re: Contribution to 15.4g FSK-PHY
    Abstract: Analysis of the performance and complexity of LDPC to the convolutional and RS codes
    being considered for FEC.
    Purpose: Contribution to the CPP merged PHY
    Notice: This document has been prepared to assist the IEEE P802.15. It is offered as a basis
    for discussion and is not binding on the contributing individual(s) or organization(s). The material
    in this document is subject to change in form and content after further study. The contributor(s)
    reserve(s) the right to add, amend or withdraw material contained herein.
    Release: The contributor acknowledges and accepts that this contribution becomes the property
    of IEEE and may be made publicly available by P802.15.

    Submission Slide 1 Slide 1


    Q. Wang [USTB]
    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    LDPC Code Performance and Complexity


    Comparing with Convolutional and RS
    Code

    Submission Slide 2 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Outline
    • Background
    • Simulation Methodology
    • Simulation Results
    – Packet Error Rate (PER) vs. SNR
    – LDPC – Convolutional coding gain difference vs. Block size
    – Impact of estimated SNR
    – Computational complexity comparison between LDPC code and RS
    code
    • Summary and Conclusions
    • References

    Submission Slide 3 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Background
    • Advanced coding candidates: BCH Code, Reed-Solomon Code, Convolutional
    Code, Turbo Code, Low Density Parity Check (LDPC) Code, etc.
    • Contribution IEEE 802.11-03/865 [1] introduced Low-Density Parity-Check
    (LDPC) codes as candidate codes for 802.11n applications. It showed
    potential advantages of those codes over existing convolutional codes used in
    802.11a/g.
    • We compare the performance of example LDPC codes with the Convolutional
    Code in 802.11n, including
    – Various frame lengths
    – Various code rates
    – Impact of estimated SNR
    • We compare the Computational Complexity of the LDPC with the RS code in
    DVB-C.
    • In this report, the performance comparison under AWGN channel is addressed
    only. In the next related submission, emphasis will be on performance
    comparison under other channel model.

    Submission Slide 4 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Simulation Methodology - General


    • PHY model with BPSK constellation. Simulation included:
    Modulation BPSK BPSK

    Coding Rate (R) 1/2 2/3

    • Channels simulated:
    – AWGN channel. This implementation utilized the MATLAB code.

    • Simulation scenario assumed:


    – All packets detected, ideal synchronization, no frequency offset
    – Ideal front end, Nyquist sampling frequency

    Submission Slide 5 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Simulation Methodology - FEC


    • General FEC:
    – Code lengths: 648, 1296 bits, chosen based on 802.11n standard [2]
    – Code rates: 1/2, 2/3 (as in 802.11n)

    • Convolutional codes:
    – Viterbi decoding algorithm

    • LDPC codes:
    – Iterative Sum-Product decoding algorithm (BP) with 20 iterations
    – Concatenated codewords for longer packets

    Submission Slide 6 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Simulation Results: PER vs. SNR


    Channel Model: AWGN
    Modulation: BPSK
    0
    Packet Error Rate (PER) vs. SNR under AWGN Channel
    10

    -1
    10

    -2
    10
    PER

    -3
    10

    LDPC Code with packet length 648, coding rate 1/2


    LDPC Code with packet length 648, coding rate 2/3
    -4
    LDPC Code with packet length 1296, coding rate 1/2
    10 LDPC Code with packet length 1296, coding rate 2/3
    Convolutional Code with packet length 648, coding rate 1/2
    Convolutional Code with packet length 648, coding rate 2/3
    Convolutional Code with packet length 1296, coding rate 1/2
    Convolutional Code with packet length 1296, coding rate 2/3
    -5
    10
    0 1 2 3 4 5 6 7 8 9 10
    SNR (dB)

    Submission Slide 7 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Simulation Results:
    (LDPC_coding_gain– Convolutional_Coding_Gain) vs. Block Size
    Modulation: BPSK
    Code rate: 1/2
    Channel model: AWGN
    Coding gain difference measured at PER of 10-2
    LDPC-Convolutional Coding gain difference at PER of 10-2 vs. block size
    6.5

    6
    C o d in g g a in d iff e re n c e (d B )

    5.5

    4.5

    3.5

    2.5
    0 200 400 600 800 1000 1200 1400
    Block size (bits)

    Submission Slide 8 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Simulation Results: Impact of estimated SNR

    Impact of estimated SNR x-axis indicates:


    0
    10
    estimated _ SNR
      10  log10
    with ideal SNR Estimator
    ideal _ SNR
    with non-ideal SNR Estimator
    -1
    10 Where ideal_SNR denotes the
    variable used to generate AWGN
    and estimated_SNR denotes the
    -2
    variable got by SNR estimation
    BER

    10
    algorithm.
    Modulation type: BPSK
    -3
    10
    Coding rate: 1/2

    Code length after encoder: 1944


    -4
    10
    -2 0 2 4 6 8 10 12 14 16
    10log10(Est-SNR/Ideal-SNR)

    Submission Slide 9 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Complexity Comparison between LDPC Code and RS Code


    For both LDPC decoder and RS decoder, the implementation complexity heavily
    depends on the decoding algorithm, e.g. BP/log-BP/min-sum for LDPC and
    architecture & logic design. Thus, we only discuss the computational complexity in
    terms of big-O.

    • According to reference [3], the decoding complexity for one iteration of the
    BP decoding is:
    Addition operation: N (3J  1)
    Multiplication operation: 4 N (3 J  2)
    where N is the code length of LDPC code; J is the number of ones in each
    column.
    • According to reference [4], the decoding complexity for RS decoding is:
    Addition and multiplication operation in Galois Field:
    2t (n  1)  2tu  nu  2u 2
    where t is correct ability of RS code; n is code length; u is the number of
    errors for one packet.

    Conclusion: The computational complexity of LDPC Code increases


    linearly with incensement of block size as that of RS Code.
    Submission Slide 10 Q. Wang [USTB]
    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    Summary and Conclusions

    • LDPC codes offer considerable performance advantages over the existing


    convolutional codes.

    • With the proper design LDPC codes can be made flexible enough in terms of
    coding rate and block size, so as to satisfy demands of 802.15.4g applications.

    • The decoding algorithm of LDPC presented here is not sensitive to the accuracy
    of SNR estimating.

    • The computational complexity of LDPC Code increases linearly with


    incensement of block size as that of RS Code.

    Submission Slide 11 Q. Wang [USTB]


    Sept 2009 doc.: IEEE 802. 15-10-0241-00-004g

    References
    [1] IEEE 802.11-03/865r1, “LDPC FEC for IEEE 802.11n Applications”, Eric
    Jacobson, Intel, November 2003.
    [2] IEEE Std 802.11n/D2.00, Part 11: Wireless LAN Medium Access Control
    (MAC) and Physical Layer (PHY) Specifications, Enhancements for Higher
    Throughput.
    [3] Marc P. C. Fossorier, Miodrag Mihaljevic, “Reduced Complexity Iterative
    Decoding of Low-Density Parity Check Codes Based on Belief Propagation”,
    IEEE Transactions on Communications, Vol. 47, No. 5, May, 1999.
    [4] Hao Yongjie, Jiang jianguo, “Improved Time-domain Decoding Algorithm of
    RS Code”, Computer Engineering, Vol. 34, No. 14, July 2008.

    Submission Slide 12 Q. Wang [USTB]

    You might also like