Traditionally, computer storage has been separated into a hierarchy based on response time, volatility, and cost of media. This tiering is undergoing a significant upheaval as a new breed of memory technologies, termed Storage Class Memories (SCM), now make it feasible to replace several tiers of the hierarchy with a single, cost-effective, uniform type of memory/storage. To make large-scale SCM deployments practical, however, memory system designers will first need to solve the problem of how to guard against unavoidable storage wear-out and failures-problems traditionally absent from "main memory" and handled by software at leisurely timescales in the domain of storage. In this paper, we propose a novel approach to providing fault tolerance in SCM-based main memory. Our key insight is to treat memory as a distributed storage system and rely on data replication and a consensus protocol to keep the replicas consistent. Separate memory instances store replicated copies of the data, and we use a programmable network interconnect to provide fast consensus between the memory instances. Our initial experiments using software memory controller emulation demonstrate reasonable overhead over local memory reads and show great promise as scalable main memory.

Industrial applications and cyber-physical systems rely on real-time wireless networks to deliver data in a timely and reliable manner. However, existing solutions provide these guarantees only for stationary nodes. In this article, we present REWIMO, a solution for real-time and reliable communications in mobile networks. REWIMO has a two-tier architecture composed of (i) infrastructure nodes and (ii) mobile nodes that associate with infrastructure nodes as they move. REWIMO employs an on-join bandwidth reservation approach and benefits from a set of techniques to efficiently reserve bandwidth for each mobile node at the time of its admission and over its potential data forwarding paths. To ensure association of mobile nodes with infrastructure nodes over high-quality links, REWIMO uses the two-phase scheduling technique to coordinate neighbor discovery with data transmission. To mitigate the overhead of handling network dynamics, REWIMO employs an additive scheduling algorithm, which is capable of additive bandwidth reservation without modifying existing schedules. Compared to the algorithms used by static real-time wireless networks, the techniques and the algorithms employed by REWIMO result in a significant increase in real-time capacity, enhanced reliability, and considerably faster handling of network dynamics.

Real-time wireless communication infrastructure is increasingly deployed to support industrial and cyber-physical applications. A limitation of existing real-time protocols is that they do not support mobility. This paper presents the development of a real-time network composed of a multi-hop infrastructure, and mobile nodes that associate with infrastructure nodes as they move. Once a mobile node joined the network, its real-time communication is guaranteed irrespective to the number and mobility pattern of mobile nodes. To develop this network, we propose Mobility-Aware Scheduling Algorithm (MASA), which benefits from new transmission scheduling approaches that cleverly combine potential packet transmissions to increase real-time capacity. We have developed a realistic trace-based simulator to evaluate the performance of MASA against two baseline algorithms. Experimental results indicate that MASA increases the number of admitted mobile nodes by 7× and 1.6×, and extends the network lifetime by 110% and 30%, compared to the baselines.

In this paper we consider the problem of supporting real-time communication in mobile networks. To address this challenge, we propose novel transmission scheduling techniques that handle the routing uncertainty introduced by mobility. The core of the scheduling techniques involves controlling the order in which transmissions are scheduled and intelligently scheduling multiple transmissions in a single entry of the scheduling matrix without conflict. Flow-Ordered Mobility-Aware Real-time Scheduling (FO-MARS) integrates these techniques to provide a 14x increase in real-time capacity compared to the baseline algorithms designed for static real-time networks. Additionally, we propose Additive Mobility-Aware Real-time Scheduling (A-MARS), which can handle network dynamics such as the addition or removal of flows without having to reschedule previously admitted flows. As a result, A-MARS achieves significantly lower admission latency than that of the baselines.

Energy-efficiency is a critical concern in continuously-running mobile applications, such as those for health and context monitoring. An attractive approach to saving energy in such applications is to defer the execution of delay-tolerant operations until a time when they would consume less energy. However, introducing delays to save power may have a detrimental impact on the user experience. To address this problem, we present Tempus, a new approach to managing the trade-off between energy savings and delay. Tempus saves power by enabling programmers to annotate power-hungry operations with states that specify when the operation can be executed to save energy. The impact of power management on timeliness is managed by associating delay budgets with objects that contain time-sensitive data. A static analysis and the run-time service ensure that power management policies will not delay an object more than its assigned budget. We demonstrate the expressive power of Tempus through a case study of optimizing two real-world applications. Furthermore, laboratory experiments show that Tempus may effectively manage the energy-delay trade-off on realistic workloads. For example, in a news application, five Tempus annotations may be used to create a policy that reduces the latency of downloading images 10 times compared to the original implementation without affecting energy consumption. Our experiments also indicate that the overhead of tracking budgets in Tempus is small.

The increasing growth of low-power wireless networks in real-world implementations has intensified the need to develop well-organized key network building blocks. Neighbor discovery, link quality measurement and data collection are among the fundamental building blocks of network initialization process. Over the past decade, network initialization has attracted significant attention from the research community of low-power wireless networks. Accordingly, the general concern of this paper is to survey neighbor discovery, link evaluation and collection tree construction protocols, as well as, research challenges in these research areas. Furthermore, we explore the impacts of these protocols on the functionality of different layers in the network protocol stack. In order to provide a clear view of the state-of-the-art neighbor discovery approaches, this paper also presents a classification of the existing neighbor discovery protocols. Finally, some of the important open issues in developing network initialization protocols are discussed to present new directions for further research.