Recursive Pseudo-Bayesian Access Class Barring for M2M Communications in LTE Systems
Recursive Pseudo-Bayesian Access Class Barring for M2M Communications in LTE Systems
复制标题
DOI：
10.1109/tvt.2017.2681206
复制
发表时间：
2017-03
影响因子：
6.8
通讯作者：
Hu Jin;Waqas Tariq Toor;B. Jung;Jun-Bae Seo
中科院分区：
文献类型：
--
作者：
Hu Jin;Waqas Tariq Toor;B. Jung;Jun-Bae Seo
商用长期演进（LTE）系统在初始随机接入过程中采用多前导接入类限制（ACB）机制，以适应机器类型通信的突发流量到达。在本文中，我们提出了两种贝叶斯ACB算法，这两种算法仅基于每个插槽中空闲序文的数量来估计活跃机器设备的数量。在商用LTE系统中，eNodeB无法立即区分特定的前文是来自单个设备（即成功）还是多个设备（即碰撞）。但是，空闲的前奏可以在每个时隙的基站（BS）上立即检测到。数值结果表明，在已知有源设备数目的情况下，本文提出的算法与理想ACB算法的性能相当。
Commercial long-term evolution (LTE) systems adopt an access class barring (ACB) mechanism in the initial random access procedure with multiple preambles in order to accommodate bursty traffic arrivals of machine-type communications. In this paper, we propose two Bayesian ACB algorithms that estimate the number of active machine devices based only on the number of idle preambles in each slot. In the commercial LTE systems, eNodeB cannot instantaneously distinguish if a particular preamble is sent from a single device (i.e., success) or multiple devices (i.e., collision). However, the idle preambles can be instantaneously detected at the base station (BS) in each slot. Numerical results show that the proposed algorithms yield quite similar performance with the ideal ACB algorithm, assuming that the exact number of active devices is known to the eNodeB.