Penglin Dai

Vehicle Assisted Data Update for Temporal Information Service in Vehicular Networks

Vehicular networks blueprint the bright future of transportation systems in safety, efficiency and sustainability. Highly dynamic traffic information is one of the most important features in vehicular networks, which makes data services very challenging, as the data quality drops over time dramatically and timely data update is expected to maintain the service quality. In this paper, we consider the system architecture, in which vehicles coming from different directions are able to sense and carry upto-date location-based information along their trajectories and upload fresh information to the roadside unit (RSU) when passing through. Meanwhile, vehicles may request information from the RSU for other routes. To enable efficient data services in such a scenario, firstly, we characterize the freshness of temporal data. Then, based on the general form of data quality function, we propose a heuristic algorithm called priority-based scheduling (PBS), which synthesizes the data quality, the broadcast effect and the real-time service requirement in making scheduling decisions. A comprehensive performance evaluation demonstrates the superiorty of PBS under a variety of scenarios.

Write reconstruction for write throughput improvement on MLC PCM based main memory

The emerging Phase Change Memory (PCM) is considered as one of the most promising candidates to replace DRAM as main memory due to its better scalability and non-volatility. With multi-bit storage capability, Multiple-Level-Cell (MLC) PCM outperforms Single-Level-Cell (SLC) in density. However, the high write latency has been a performance bottleneck for MLC PCM for two reasons: First, MLC PCM has a much longer programming time; Second, the write latencies of different cell state transitions range significantly. When cells are concurrently written in the burst mode, the write latency of a burst is delayed by the worst state transitions. To improve the write throughput of MLC PCM based main memory, this paper proposes a Write Reconstruction (WR) scheme. WR reconstructs multiple burst writes targeting the same memory row, where the worst case cells are grouped together at some writes. With this approach, the write latency of other writes will be reduced. WR incurs low implementation overhead and shows significant efficiency. Experimental results show that WR achieves 18.1% of write latency reduction on average, with negligible power overhead.

Effective file data-block placement for different types of page cache on hybrid main memory architectures

Hybrid main memory architectures employing both DRAM and non-volatile memories (NVMs) are becoming increasingly attractive due to the opportunities for exploring benefits of various memory technologies, for example, high speed writes on DRAM and low stand-by power consumption on NVMs. File data-block placement (FDP) on different types of page cache is one of the important problems that directly impact the performance and cost of file operations on a hybrid main memory architecture. Page cache is widely used in modern operating systems to expedite file I/O by mapping disk-backed file data-blocks in main memory to process space in virtual memory. In a hybrid main memory, different types of memory with different read/write costs can be allocated as page cache by operating system. In this paper, we study the problem of file data-block placement on different types of page cache to minimize the total cost of file accesses in a program. We propose a dynamic programming algorithm, the FDP Algorithm, to solve the problem optimally for simple programs. We develop an ILP model for the file data-block placement problem for programs composed of multiple regions with data dependencies. An efficient heuristic, the global file data-block placement (GFDP) Algorithm, is proposed to obtain near-optimal solutions for the problem of global file data-block placement on hybrid main memory. Experiments on a set of benchmarks show the effectiveness of the GFDP algorithm compared with a greedy strategy and the ILP. Experimental results show that the GFDP algorithm reduces the total cost of file accesses by

Towards Real-Time and Temporal Information Services in Vehicular Networks via Multi-Objective Optimization

51.3~\%

Optimizing data placement for reducing shift operations on domain wall memories

51.3% on average compared with the the greedy strategy.

Quality-of-Experience-Oriented Autonomous Intersection Control in Vehicular Networks

This paper embarks the first study on exploiting the synergy between vehicular caching and network coding for enhancing the bandwidth efficiency of data broadcasting in heterogeneous vehicular networks by presenting a service architecture that exercises the software defined network concept. In particular, we consider the scenario where vehicles request a set of information and they could be served via heterogeneous wireless interfaces, such as roadside units and base stations (BSs). We formulate a novel problem of coding-assisted broadcast scheduling (CBS), aiming at maximizing the broadcast efficiency for the limited BS bandwidth by exploring the synergistic effect between vehicular caching and network coding. We prove the NP-hardness of the CBS problem by constructing a polynomial-time reduction from the simultaneous matrix completion problem. To efficiently solve the CBS problem, we employ memetic computing, which is a nature inspired computational paradigm for tackling complex problems. Specifically, we propose a memetic algorithm, which consists of a binary vector representation for encoding solutions, a fitness function for solution evaluation, a set of operators for offspring generation, a local search method for solution enhancement, and a repair operator for fixing infeasible solutions. Finally, we build the simulation model and give a comprehensive performance evaluation to demonstrate the superiority of the proposed solution.

Adaptive scheduling for real-time and temporal information services in vehicular networks

Real-time and temporal information services are intrinsic characteristics in vehicular networks, where the timeliness of data dissemination and the maintenance of data quality interplay with each other and influence overall system performance. In this work, we present the system architecture where multiple road side units (RSUs) are cooperated to provide information services, and the vehicles can upload up-to-date information to RSUs via vehicle-to-infrastructure (V2I) communication. On this basis, we formulate the distributed temporal data management (DTDM) problem as a two-objective problem, which aims to enhance overall system performance on both the service quality and the service ratio simultaneously. Further, we propose a multiobjective evolutionary algorithm called MO-DTDM to obtain a set of pareto solutions and analyze how to fulfill given requirements on system performance with obtained pareto solutions. Finally, we build the simulation model and give a comprehensive performance evaluation, which demonstrates the superiority of the proposed optimization method.

A Convex Optimization Based Autonomous Intersection Control Strategy in Vehicular Cyber-Physical Systems

As a special yet important part of vehicular ad hoc networks (VANETs), Bus-VANETs have attracted much research attention in recent years. Most of existing studies focused on developing routing protocols in Bus-VANETs based on the predictable bus route. In this work, we embark a study on information services in Bus-VANETs, where bus information is provided via the hybrid of infrastructure-to-vehicle (I2V) and vehicle-to-vehicle (V2V) communications. In particular, we consider the scenario where buses on different routes request different information from roadside units (RSUs), which are the static infrastructures installed along the road, and provide information services to passing buses via I2V communication. Furthermore, buses are also able to share information with each other via V2V communications. On this basis, we formulate the cooperative data sharing problem by considering the requested information and the encounter probability of buses. Then, we put forward an offline Integer Linear Programming (ILP) solution based on the knowledge of bus traces to obtain the optimal performance. Moreover, we propose an on-line Prediction-based Data Sharing (PDS) algorithm by analyzing the encounter probability of buses. Based on the analysis, each RSU computes the utility of broadcasting data items and makes the scheduling decision with the cooperation of other RSUs. Finally, we build the simulation model and implement the proposed solution for performance comparison. The experiment results demonstrate that PDS can achieve near-optimal performance with much lower computational overhead, and it can outperform other alternative on-line scheduling algorithms significantly.