A.R.Q. techniques using Sub-block retransmission for wireless networks

Document Sample
A.R.Q. techniques using Sub-block retransmission for wireless networks Powered By Docstoc
					                                                   (IJCSIS) International Journal of Computer Science and Information Security,
                                                   Vol. 8, No. 6, September 2010

            A.R.Q. techniques using Sub-block retransmission for wireless networks

                A.N.Kemkar, Member, ISTE and Dr. T.R.Sontakke Member,ISTE

Abstract—In this paper we mainly focus our                           In an attempt to improve throughput
investigation on the throughput performance in                    performance, we have included an analysis using
conjugation with sub-block transmission scheme.                   forward error correcting (FEC) block codes (used
The     throughput    of     a   wireless   data                  in Hybrid ARQ). The optimum amount of FEC
communications system depends on a number of                      coding was found to be dependent upon the Block
variables, one of it is length of the message                     length. As the Block length increases, the number
blocks. Over a noisy communication medium like                    of correctable errors to optimize the throughput
wireless medium used for mobile ad-hoc network,                   also increases, mathematical expression is shown
our propose scheme performs effectively. In                       in 2.2.
propose scheme random length of the message is                    The paper is organized as follows. In Section 2
divided in to fixed length blocks and applying                    .Summary on the Related work and basic concept.
ARQ techniques if the error occurs. A threshold                   In section 3.scheme description and system
model is used for fading channel, estimation and                  model. We consider the performance analysis of
CRC detection codes are used. Comparison of                       the proposed scheme for simulation study in
transmission efficiency of proposed scheme with                   section 4. followed by conclusion in Section 5
varying channel condition is shown.     .                         2.RELATED WORK :
.Index Terms— FEC, Hybrid ARQ, BER.                               2.1: Related Work- The efficiency of HARQ
                                                                  scheme is compared with GBN schemes using
1. INTRODUCTION:                                                  different lengths of IP Blocks. Further show that
Wireless channels are highly affected by                          usage of smaller Blocks and hybrid schemes leads
unpredictable factors such as co-channel                          to an improved throughput. Differences between
interference, adjacent channel interference,                      pure and hybrid GBN schemes are also discussed.
propagation path loss, shadowing and multi path                   [1]
fading. The unreliability of media degrades the                   When the channel is quiet the sub-block
transmission quality seriously. Automatic Repeat                  retransmission     scheme     behaves     like  a
ReQuest (ARQ) and Forward Error Correction                        conventional ARQ or hybrid ARQ scheme. As
(FEC) schemes are frequently used in wireless                     the channel becomes increasingly noisy, the data
environments to reduce the high bit error rate of                 block is divided into smaller sub-blocks for
the channel.                                                      transmission. Each sub-block is encoded for error
As we have seen, the throughput efficiencies of                   control by an appropriate shortened code of
all the basic ARQ schemes are functions of the                    which the code length is adapted to the
packet size n. [1],[2],[3],[4],[5]. Our main result               corresponding channel BER.[2] Further optimum
is a mathematical technique for determining the                   block size in accordance with the channel
block size as a function of the other variables like              conditions [4] A single code HARQ scheme was
BER, signal-to-noise ratio.                                       proposed in which transmitter is operating in any
--------------------------------------------------------          one mode with the degree of errors encounter.
A.N.Kemkar1 ,S.R.T.M.U,Nanded.                                    The operating state is selected based on the
+91-9819150392,ankemkar@gmail.com                                 channel BER. Data bits are divided in blocks and
Dr.T.R.Sontakke2                                                  are encoded with shortened codes. During the
                                                                  retransmission new coded blocks are combined
Ex.Director – S.G.G.S.I.T.E.- Nanded
                                                                  and at the receiver end proper decoding
Principal,Sidhant college of Engineering                          techniques are used to separate retransmitted
Pune.+91-                                                         blocks from the new blocks.
9822392766,trsontakke@gmail.com                                   2.1. BASIC CONCEPT -

                                                            234                               http://sites.google.com/site/ijcsis/
                                                                                              ISSN 1947-5500
                                                          (IJCSIS) International Journal of Computer Science and Information Security,
                                                          Vol. 8, No. 6, September 2010

Analytical     expression     how     Throughput                               L-C
                                                                                   R( 1− P e( γ ) )
                                                                          T=                                (5)
performance of the system varies with the size of                               L
the block length and FEC. :Consider the
following two cases to verify the Throughput                             Case 2.: Now instead of transmitting those L bits
performance of the system.                                               with no error correction capability, we will now
Case 1. Blocks are transmitted without FEC.                              add B error correcting bits and transmit a total of
Case 2. Blocks are transmitted with FEC.                                 bits L+B .Using a block code forward error
Our analysis includes the following simplifying                          correction scheme, the minimum number of B
assumptions:1. The CRC decoder detects all                               bits required to correct t errors is given by [5]
errors in the output of the FEC decoder.2.                                           t  L +B  
Transmission of acknowledgments from the                                  B ≥ log 2  ∑                            (6)
                                                                                     n =0  n  
                                                                                             
receiver to the transmitter is error free and
instantaneous.                                                           Now that we can correct‘ t ’ errors, our block
System throughput ( T ) is the number of payload                         success rate, f (γ ) should be larger than its
bits per second received correctly:                                      previous value with no error correction. Recall
                                                                         that   f (γ )   with t = 0 is given by:
T=   Rf (γ )   (1)                                                                                     L
   L                                                                      f (γ ) = ( 1 − P e ( γ ) )       where    Pe ( γ )   is the probability
       KR 
where   b/s is the payload                 transmission rate           of a bit error as a function of the SNR. Now, with
       L                                                               error correction capability, the Block success rate
Where f (γ ) = Block success      rate defined as                        for some arbitrary value of t is [7]
probability of receiving block correctly.                                            t  L +B                            L +B -n
Probability is a function of signal to noise ratio.                       f t(γ ) = ∑        P n ( γ ) ( 1 − Pe ( γ ) )                    (7)
                                                                                         n  e
  E                                                                                n =0      
γ= b               (2)
  No                                                                     Our new equation for the throughput as a function
                        P                                                of the signal to noise ratio is:
In which E b = joules in received energy per bit.                                          P 
                                                                                              
where R = Transmission rate. Probability is a                                     L-C   N o 
                                                                          T(γ ) =              f t(γ )                   (8)
function of signal to noise ratio.                                                L +B   γ 
                                                                                              
      Eb   P                                                                                  
γ =      =                       (3)
      N o N oR
                                                                         From (5) and (8) it is clear that throughput of the
Each Block, of length L bits, is a combination                           system is a function of message block length.
of a payload ( K ) and overhead ( L-K ) . Because                        Further (5) and (8) are used for pure and hybrid
the Block success rate, f (γ ) is a decreasing                           ARQ techniques.
function of, L there is an optimum Block length,.
 L • . When
                                                                         3. SCHEME DISCRIPTION AND SYSTEM
 L p L∗ excessive overhead in each Block limits                          MODEL :
                                                                         This paper presents a sub-block retransmission
the throughput. When L > L ∗ Block errors limits
                                                                         scheme for ARQ . The data block is divided into
the throughput.
                                                                         smaller sub-blocks for transmission. Each sub-
For case 1. When there is no forward error
                                                                         block is encoded by an appropriate error detection
correction coding. In this case
                                                                         codes. The encoded block is then transmitted. The
f (γ ) = ( 1 − P e ( γ ) )             (4)                               received block is checked for errors sub-block by
where f (γ ) = block success rate defined as                             sub-block. The proposed scheme provides
probability of receiving block correctly.                                improved throughput by retransmitting only the
 P e ( γ ) is block error rate.                                          sub-blocks in the occurrence of errors.
Therefore, in a system without FEC, the
                                                                         3.1 SYSTEM MODEL :
throughput as a function of L ,from (1)

                                                                   235                                       http://sites.google.com/site/ijcsis/
                                                                                                             ISSN 1947-5500
                                               (IJCSIS) International Journal of Computer Science and Information Security,
                                               Vol. 8, No. 6, September 2010

  We consider an ad-hoc network with V nodes                  SIMULATION RESULTS: We evaluate the
and assume that each node is equipped with only               performance     of     the    proposed      scheme
one antenna. A Point to Point protocol is used at             implemented with Matlab. We run the simulation
the medium access control layer. A Selective                  for two schemes i.e. with sub block transmission
repeat request ARQ mechanism is used.                         and without sub block transmission. The
Particularly, the source node transmits a data                simulation parameters are shown in the table
packet with a C-bit CRC attached. The                         1.Simulation run for 5000 total blocks. Result is
destination node detects CRC and then sends an                the average of independent experiments where
acknowledgement that is either positive (ACK) or              each experiment uses different randomly
negative (NACK) back to the source node. If the               generated uniform parameters. We use mean
packet is correctly detected by the destination               values which are obtained independent
node (with ACK feedback), the source node                     experiments as a basic data to get the result.
continues to transmit a new data packet and the               Simulation results are shown Table 2
above     process   is     repeated.   Otherwise,              Table 1: System Parameters:
retransmission will start. A threshold model for              Parameters                         Notaion         Values
channel characterization is used for fading                   Signal to Noise Ratio                   γ          Varied
channel.                                                      Total number of blocks             -------         5000
4. PERFORMANCE ANALYSIS OF THE                                Total sub block                     M              32
PROPOSE SCHEME:                                               Information bits in a block         K              16
                                                              Packet length                       n              5000*32*16
The performance analysis of the scheme is
                                                              Max.         number        of      --------        3
measured in terms of throughput of the proposed               Retransmissions
scheme. Further we show the comparison of                     Number of sub blocks                L              Varied
throughput with sub block and without sub block               retransmitted
transmission schemes.                                         Cyclic Redundancy Check            CRC             Varied
Expression of throughput for ARQ for present                  Bit error rate                     BER             Varied
                                                              Packet error rate                  PER             Varied
scheme:                                                                                          η
                                                              Throughput efficiency                              Varied
η=              (9)
     E[T]                                                     Table 2 : Simulation Results: Following
where K =information bits in a block. E [ .]                  Simulation results shows the comparison of
=Expectation of number of transmitted bits in a               Throughput efficiency verses varied block size
given block.                                                  verses changing channel condition in terms of
           ∞                                                  PER.
T = Mn + ∑ T i           (10)                                 Table 2:
           i =1
 where M =number of sub blocks, n =number of                  Block length               Packet error rate       Throughput
bits in a sub block, T i =number of transmitted bits          Without      With          Without With            Without With
                                                              sub block    sub           sub        sub          sub      sub
for i th transmission.                                                     block         block      block        block    block
E [ T ] = Mn + ∑ E [ T i ] (11)
                  i =1                                        Whole           Whole      0.1          0.1        0.9             0.99
where E [ T ] =Average number of transmitted                  block     is
                                                                                         0.3          0.3        0.86            0.96
                                                                                         0.5          0.5        0.66            0.94
bits.                                                         with out        divided    0.7          0.7        0.59            0.93
Out of M sub blocks if L sub blocks are                       sub             into       0.9          0.9        0.57            0.9
                                                              division.       sub 4      1            1          0.5             0.89
transmitted at the i th retransmission ,then random                           blocks
variable, T i takes the value Ln ,if L out of M
sub-blocks are retransmitted at the i th

                                                        236                               http://sites.google.com/site/ijcsis/
                                                                                          ISSN 1947-5500
                                                                                            (IJCSIS) International Journal of Computer Science and Information Security,
                                                                                            Vol. 8, No. 6, September 2010

From the Fig.1 shown below performance of the                                                              6. REFERENCES:
proposed scheme. This paper has presented a sub-
block retransmission scheme for                                                                            [1]       W.W. Chu. Optimalmessage block size
1. With micro block ARQ                                                                                              for computer communications with error
2. Without micro block ARQ                                                                                           detection and retransmission strategies.
                                                                                                                     IEEE Transactions on Communications,
Proposed sub-block retransmission schemes                                                                            COM-22:1516– 1525, October 1974.
showed better overall performance compared to                                                               [2]        J.S. DaSilva, H.M. Hafez, and S.A.
the other competitive schemes by retransmitting                                                                      Mahmoud. Optimal packet length for
without sub-blocks in the occurrence of errors.                                                                      fading land mobile data channels. In
From Table 2 and the Fig.1 it is clear that as the                                                                   Proceedings of ICC 1980, pages 61.3.1–
packet error rate increases i.e. channel condition                                                                   61.3.5, June 1980.
gets deteriorated the throughput performance
goes down by more than 40% as compare to                                                                   [3]       R.L. Kirlin. Variable block length and
proposed scheme. There fore we proposed sub-                                                                         transmission      efficiency.     IEEE
block retransmission scheme with ARQ is more                                                                         Transactions on          Communication
reliable than existing i.e. without sub block                                                                        Technology, COM-17:350–355, June
transmission scheme.                                                                                                 1969.

Fig.1: Performance of proposed scheme vs.                                                                   [4]      E. Modiano. An adaptive algorithm for
existing schemes.                                                                                                    optimizing the packet size used in
                                      T h ro u g h p u t V s . P a c k e t E rro r R a te                            wireless ARQ protocols. Wireless
                                                                                                                     Networks, 5:279–286, July 1999.
               0 .9 5

                0 .9                                                                                        [5]      J.M.Morris. Optimal block lengths for
                                                                                                                     ARQ error      control schemes. IEEE
               0 .8 5
                                                                                                                     Transactions on Communications, COM-
                0 .8                                              W it h M ic ro b lo c k                            27:488–493, February 1979
                                                                  W it h o u t M ic ro b lo c k
 T hroughput

               0 .7 5

                0 .7

               0 .6 5

                0 .6

               0 .5 5

                0 .5
                   0    0 .1   0 .2   0 .3   0 .4 0 .5 0 .6 0 .7 0 .8           0 .9        1
                                             P a c k e t E rro r R a te

                                                                                                     237                               http://sites.google.com/site/ijcsis/
                                                                                                                                       ISSN 1947-5500

Description: IJCSIS is an open access publishing venue for research in general computer science and information security. Target Audience: IT academics, university IT faculties; industry IT departments; government departments; the mobile industry and computing industry. Coverage includes: security infrastructures, network security: Internet security, content protection, cryptography, steganography and formal methods in information security; computer science, computer applications, multimedia systems, software, information systems, intelligent systems, web services, data mining, wireless communication, networking and technologies, innovation technology and management. The average paper acceptance rate for IJCSIS issues is kept at 25-30% with an aim to provide selective research work of quality in the areas of computer science and engineering. Thanks for your contributions in September 2010 issue and we are grateful to the experienced team of reviewers for providing valuable comments.