Random Access for M2M Communications over LTE network

1. Random Access for M2M Communications over LTE network

2. Outline • Problem Statement • Related work • Proposed Design • References

3. Problem Statement • To enable full mechanical automation where each smart device can play multiple roles among sensor, decision maker, and action executor, it is essential to construct scrupulous connections among all devices [1] • In the literature, few realizations of M2M communications have been proposed, such as leveraging Bluetooth (IEEE 802.15.1), Zigbee (IEEE 802.15.4), or WiFi (IEEE 802.11b) technology. However, there is still no consensus on the network architecture of a general scenario for M2M communications [1]

4. Problem Statement • It is one of the key issues in the 4-th generation (4G) cellular networks how to efficiently handle the heavy random access (RA) load caused by newly accommodating the huge population of Machine-to-Machine or Machine-Type Communication (M2M or MTC) customers/devices [2] • Supporting trillions of devices is the critical challenge in machine-to-machine (M2M) communications, which results in severe congestions in random access channels of cellular systems that have been recognized as promising scenarios enabling M2M communications [3]

5. Related Work

6. Proposed Design • Problem formulation [3] • Consider the random access of M2M communications in LTE-Advanced with 𝑀 BSs indexed by 𝑚 = 1, . . .,𝑀 and 𝑁active MTC devices indexed by 𝑛 = 1, . . .,𝑁 • Distinct from the ordinary ACB that each MTC device can only access the BS attached by the MTC device, in this letter, we propose that an MTC device is able to access the BS unattached by the MTC device when the MTC device locates within the overlapped coverage area of multiple BSs

7. Proposed Design • To formulate such a random access problem, we adopt following notations • Denote 𝐴𝑚 as the set of MTC devices attaching to the𝑚th BS and ∥𝐴𝑚∥ as the norm of 𝐴𝑚. • When cooperation of BSs are available, 𝐴𝑚 for all • m are known by all BSs. Note that

8. Proposed Design • Denote 𝑀𝑛 as the set of BSs that the 𝑛th MTC device can possibly access. The 𝑛th MTC device selects one BS from 𝑀𝑛 to proceed to the random access. In LTE-Advanced, the BS can request the MTC device to perform the exploration of surrounding BSs, and report the exploration result. Thus, 𝑀𝑛 for all 𝑛 are available for all BSs. However, 𝑀𝑛 for 𝑛 = 1, . . .,𝑁, 𝑛 !=𝑗 are unknown by the 𝑗th MTC device for 𝑗 = 1, . . .,𝑁.

9. Proposed Design • Denote 𝑁𝑚 as the set of MTC devices that access the 𝑚th BS. In the ordinary ACB,∥𝑁𝑚∥for all 𝑚 are known by each BS. However, if each MTC device can access the BS unattached by the MTC device, ∥𝑁𝑚∥for all 𝑚 are random variables unknown by BSs, unless the BS selection strategy of each MTC device is given • Denote l𝑛,𝑚 as an indicator function that, for the • 𝑛th MTC device,

10. Proposed Design • Denote 𝑁′𝑚 as the set of MTC devices that can only receive the signal from the 𝑚th BS and these MTC devices can only access the 𝑚th BS. ∥𝑁′𝑚∥ is the norm of 𝑁′𝑚

11. Proposed Design

12. Proposed Design • In the ordinary ACB, although the throughput of each cell can be individually maximized by individually setting p𝑚= 1/∥𝑁𝑚∥ (where p𝑚is the ACB parameter of the the𝑚th BS) in each BS, the delay experienced by an MTC device attached to the 𝑚th BS may be unacceptable when 𝑝𝑚 requires to be set to an extremely small value under a large ∥𝑁𝑚∥ • our objective is to provide an optimum control of a set of ACB parameters p = [𝑝1, . . . , 𝑝𝑀] jointly decided by 𝑀 BSs to minimize the largest access delay experienced among 𝑁 active MTC devices

13. Proposed Design • Through cooperation among BSs to make a joint decision of p = [𝑝1, . . . , 𝑝𝑀], the problem can be formulated by

14. Proposed Design • BS selection strategy for each MTC device • BS selection strategy is based on pm, for m=1,…,M (BS with higher pm has higher probability to be accessed by MTC) • Upon receiving ˜p𝑛 = {𝑝𝑖, 𝑝𝑗, . . . , 𝑝𝑘} ⊆ p, the 𝑛th MTC device adopts the strategy • where 𝑄𝑛,𝑥 is the probability that the 𝑛th MTC device • selects the 𝑥th BS to access

15. Proposed Design • Cooperative Access Class Barring • Given that the strategy on the selection of the BS in each MTC device in (3) is known by BSs, ∥𝑁𝑚∥ for all 𝑚 can be obtained by • The optimal p can be found by algorithm 1 and algorithm 2 proposed by [3]

16. Proposed Design

17. Proposed Design

18. Proposed Design

19. 王老師的建議 • 1.如何更精確反應出每個BS下的情況? • Key paper所提的Qn,i只考量了此round期望能access BS的MTC數，逸懷的方法加入考量每台BS實際有attach上此BS的個數，更能反應出目前BS實際loading情況 • 2.Access Barring Check(ABC)與Access Class Barring(ACB)的差別? • ABC與ACB都是根據目前BS的loading來調整其threshold，所以沒有差別

20. 曾老師的建議 • 可衡量MTC與UE通過Random access的比率 • 2.MTC與UE在Random access時，是共享or獨有Resource Block(RB)? • 目前3GPP spec並沒有針對這部分做規範，但根據survey paper，目前研究都是假設MTC與UE在Random access時是共享RB • 3.UE及MTC發Random access的頻率? • UE發完Random access與BS連上後，就不再發Random access • MTC每次的上傳或是下載前，都要先跟BS發Random access

21. References • [1] Shao-Yu Lien; Kwang-Cheng Chen; Yonghua Lin; , "Toward ubiquitous massive accesses in 3GPP machine-to-machine communications," Communications Magazine, IEEE , vol.49, no.4, pp.66-74, April 2011 • [2] Ki-Dong Lee; Sang Kim; Byung Yi; , "Throughput comparison of random access methods for M2M service over LTE networks," GLOBECOM Workshops (GC Wkshps), 2011 IEEE , vol., no., pp.373-377, 5-9 Dec. 2011 • [3] Shao-Yu Lien; Tzu-HuanLiau; Ching-Yueh Kao; Kwang-Cheng Chen; , "Cooperative Access Class Barring for Machine-to-Machine Communications," Wireless Communications, IEEE Transactions on , vol.11, no.1, pp.27-32, January 2012