SCIENCE CHINA Information Sciences, Volume 60, Issue 10: 100305(2017) https://doi.org/10.1007/s11432-017-9196-0

Aggregation transmission scheme for machine type communications

More info
  • ReceivedApr 24, 2017
  • AcceptedAug 2, 2017
  • PublishedSep 1, 2017


Massive amount of small data generated by machine type communications (MTC) will pose a challengeto the future fifth generation (5G) wireless network. Since the information from or to the machine type users aggregating closely are highly correlated, the relevance of data can be excavated by big data analysis to help improve the spectral efficiency. In this paper we proposed an aggregation transmission scheme (ATS) for MTC downlink transmissions in which the transmission order of users' data packets can be adjusted according to their relevance under the delay constraints. The users having relevance will temporally share the time slots and their data are transmitted in a multicast way so that much less timeslots are needed. We propose three different algorithms, conditional random search (CRS), standard-row algorithm (SRA), and genetic algorithm (GA) to tackle the problem of transmission order adjustment. Simulation results validate the good performance of ATS and demonstrate that SRA has the lowest complexity while GA may achieve a better performance. We also analyze the impact of different delay requirements. Our work sheds light on dealing with massive MTC data traffic for future wireless communications.


This work was partially supported by Natural Science Foundation of China (Grant No. 61461136002), Key Program of National Natural Science Foundation of China (Grant No. 61631018), Fundamental Research Funds for the Central Universities, and Huawei Innovation Research Program.

  • Figure 1

    Demonstration of proposed scheme. (a) Before adjusting; (b) after adjusting.

  • Figure 2

    Flow chart for SRA.

  • Figure 3

    Example of a chromosome.

  • Figure 4

    Single point crossover.

  • Figure 5

    Performance of different algorithms with $K=8$.

  • Figure 6

    Performance of different algorithms with $M=32,~N=256$.

  • Figure 7

    Gain ratio with different degree of delay requirements.


    Algorithm 1 Standard-row algorithm

    Input: $M$, $N$, $K$

    Output: $\boldsymbol{\widetilde{I}}$, $\boldsymbol{\widetilde{D}}$

    for $m~=~1,2,\ldots,M$ do

    for $k~=~1,2,\ldots,K$ do

    Sample $B_{m}^{k}$

    Sample $d(B_{m}^{k})$

    $\boldsymbol{\widetilde{I}}=\boldsymbol{I}$, $\boldsymbol{\widetilde{D}}=\boldsymbol{D}$

    for $m~=~1,2,\ldots,M$ do

    for $k~=~1,2,\ldots,K$ do


    let $\sum_{k~=~1}^K~\boldsymbol{D}'[a][k]$ be the minimum

    while $u<M$ do

    if $u\neq~a$ then

    for $k~=~1,2,\ldots,K$ do

    for $kk~=~1,2,\ldots,K$ do

    if $\boldsymbol{I}[u][k]~==~\boldsymbol{I}[a][kk]$ then


    if $0<~interval~~\leq~\boldsymbol{D}'[u][k]$ then





    if $interval~\leq~0$, $|interval|\leq~\boldsymbol{D}'[u][k]$ then








    end while

    $\boldsymbol{\widetilde{I}}$, $\boldsymbol{\widetilde{D}}$

Copyright 2019 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有