Workshop on Massive Uncoordinated Access Protocols

The upcoming 5G wireless communication systems will not only be "4G, but faster". Instead, they will feature two new wireless modes: massive and ultra-reliable access, respectively. These modes require rather low data rates and we put them under the term lowband communication; they will support new applications, such as massive sensing, ultra-reliable vehicular links or low-latency remote control. Although the scenarios and the performance objectives of massive access are not identical with those for ultra-reliable access, there are at least two key enablers that are common for both of them: (1) efficient communication with short packets and (2) non-orthogonal protocols for uncoordinated access. This talk will first treat the problem of communication with short packets, where the data size is comparable to the size of the metadata, i.e. control information, which is not the case in broadband communication. Communication of short packets that come from a massive number of devices and/or need to meet a latency constraint requires fundamental rethinking of the packet structure and the associated communication protocols. In the second part, the talk will discuss the role of the paradigm of coded access and Successive Interference Cancellation (SIC) in lowband communications. Coded access protocols will be discussed in the context of classical random access setting with a single receiver, but also in the context of new types of multiple access problems that arise in relation to direct device-to-device (D2D) communications.

In this talk we will illuminate some aspects and challenges of 5G random access particularly in the context of "one shot" transmission and "scalability" of massive MTC traffic. It is argued that such traffic shall not be forced into the 4G bulky access procedures and that its special characteristics will make it amenable for advanced solutions. In the first part we will derive some of the main properties to be exploited further on such as limited asynchronous access, user activity, small message size, increased pilot overhead etc. We will show how these properties can be captured using the notion of "sparsity" and introduce the relevant compressed sensing (CS) theory. In the second part we will discuss some of the recent coding approaches such as ALOHA OFDM(A), Coded ALOHA etc. and related fundamental limits in this regard. We will discuss pro and cons for each of each of the schemes. Moreover we will discuss in detail how CS based signal processing can be beneficial for such solutions. In the final part we will discuss some further upcoming challenges such as low latency and "fast" security for massive MTC access.