Catálogo de publicaciones - libros

There has been an exponential growth of applications that rely on diverse types of embedded real-time end systems and devices, such as smart phones, play stations, home appliances, consumer and industrial electronics, smart sensors and actuators. These applications require diverse types of Quality-of-Service (QoS) including timeliness, dependability, security and privacy, from the end systems/devices which are usually networked together via heterogeneous networking technologies and procotols.

We now know how to guarantee timeliness and, to a lesser extent, how to provide fault-tolerance, on both end systems and their interconnection networks. However, how to secure them is far less known, despite the growing importance of protecting information stored in the end systems/devices and exchanged over their interconnection networks. Morever, timeliness, fault-tolerance, security and privacy—which I will call simply QoS—must be supported simultaneously, often with a tight resource budget such as memory, computation and communication bandwidth, and battery power. Also, different applications require different combinations of QoS components, and hence, one-fits-all solutions are not acceptable.

This talk will start with generic aspects of QoS and then detail how to secure a sensor network for surveillance applications. Sensor networks, usually built with a large number of small, low-cost sensor devices, are characterized by their large-scale and unattended deployment that invites many critical attacks, thereby necessitating high-level security support for their intended applications and services. However, making sensor networks secure is challenging due mainly to the fact that sensors are battery-powered and it is often very difficult to change or recharge their batteries. To meet this challenge, we have been developing Lightweight Security Protocols (LiSP) that cooperatively build a unified, energy-efficient security framework for sensor networks.

Digitalization has become the unavoidable trend all around the world. The digital future holds infinite possibilities. The obvious questions are who will win and who can survive in this fast-changing world, not only in devices but also in services and contents.

Digital technology is the key locomotive that promotes the development of digital device, services and contents. The power of digital technology is getting stronger everyday in this digital era. For example, digital devices and services have merely been capable of responding to demanded requests. But now we are entering an era where the digital technology is capable of actively getting into our life, getting things done smoothly and silently. New forms of human-computer interface, such as wearable computers that are communicating with innumerable sensors deployed almost everywhere, will be common in the very near future. As ubiquitous computing evolves into an essential component of our daily lives, we need to address a variety of issues regarding social infrastructure, legal and political problems, etc. as well as technological capability.

The world has witnessed Korea turning what sounded like science fiction into everyday life. Korea is the world leader in applying new digital technologies and creating added value from them. To seize this unique opportunity for Korea, technological and social consensus among government, industry, academia, and other social bodies are needed. What are our technological strategies in developing new value-creating opportunities? What should be the role of the government in responding to the emerging digital trends? I hope that EUC will provides new insights into these challenging questions.

In most wireless sensor networks, applications submit their requests as queries and wireless sensor network transmits the requested data to the applications. However, most existing work in this area focuses on data aggregation, not much attention has been paid to query aggregation. For many applications, especially ones with high query rates, query aggregation is very important. In this paper, we design an effective query aggregation algorithm SAQA to reduce the number of duplicate/overlapping queries and save overall energy consumption in the wireless sensor networks. This new aggregation algorithm focuses on the duplicate/overlapping spatial and attribute information in the original queries submitted by the applications. Our performance evaluations show that by applying our query aggregation algorithm, the overall energy consumption can be significantly reduced and the sensor network lifetime can be prolonged correspondingly.

Energy and delay are critical issues for wireless sensor networks since most sensors are equipped with non-rechargeable batteries that have limited lifetime. Due to the uncertainties in execution time of some tasks, this paper models each varied execution time as a probabilistic random variable and incorporating applications’ performance requirements to solve the MAP (Mode Assignment with Probability) problem. Using probabilistic design, we propose an optimal algorithm to minimize the total energy consumption while satisfying the timing constraint with a guaranteed confidence probability. The experimental results show that our approach achieves significant energy saving than previous work. For example, our algorithm achieves an average improvement of 32.6% on total energy consumption.

This paper presents an innovative DVS technique to reduce the energy dissipation. Our objective is to minimize the transitions between power modes by maximizing the idle periods of functional units with instruction scheduling. Our work first analyzes the control flow graph of the application, which contains many regions. Second, we collect the power information and build its power model for each region. Then two regions with the same functional units will be merged if no dependencies exist between them. The process is repeated until no further mergings can be performed. Next, the idle functional units will be turned off and each region will be assigned a power mode based on the power model. Finally, the application is rescheduled to merge the regions to reduce the transitions between power modes. The experimental results show that our work can save the energy by 26%.

Energy-aware task scheduling is critical for real-time embedded systems. Although dynamic power has traditionally been a primary source of processor power consumption, leakage power is becoming increasingly important. In this paper, we present two optimal energy-aware polynomial-time algorithms for scheduling a set of tasks with release times, deadlines and precedence constraints on a single processor with continuous voltages. Our algorithms are guaranteed to minimise the total energy consumption of all tasks while minimising their maximum lateness under two power models: the dynamic power model where the dynamic power dominates the processor power consumption and the dynamic and leakage power model where both dynamic power and leakage power are significant sources of the processor power consumption. The time complexities of both algorithms are () , where is the number of tasks.

This paper proposes a novel power-aware scheme of H.264/AVC video player for the PAC SoC platform based on its modern dual-core architecture with DVFS capability. Energy/power is saved by the global view of power state transitions on the dual-core subsystem according to a user’s behaviors of playing a video. When the user stays in continuous video decoding, a fine-grain power-aware scheme is devised to save the energy/power in advance. Especially the fine-grain model is suitable for any standard coded H.264/AVC video without extra modifications. We also discuss a workable reduction technique when imprecise video decoding time is permitted under soft real-time constraint. For a similar SoC platform with the dual-core architecture and DVFS capability, the idea presented here is, to the best of our knowledge, the first power-aware design of H.264/AVC video player.

Multicore processors promise higher throughput at lower power consumption than single core processors. Thus in the near future they will be widely used in hard real-time systems as the performance requirements are increasing. Though DVS may reduce power consumption for hard real time applications on single core processors, it introduces a new implication for multicore systems since all the cores in a chip should run at the same performance. Blind adoption of existing DVS algorithms may result in waste of energy since a core which requires low performance should run at the same high frequency with other cores. Based on the existing partitioning algorithms for the multiprocessor hard real-time scheduling, this article presents dynamic task repartitioning algorithm that balances task loads among cores to avoid the phenomena dynamically during execution. Simulation results show that in general cases our scheme makes additional energy saving more than 10% than that without our scheme even when the schedules are generated by WFD partitioning algorithm which is known as the best energy efficient partitioning algorithm.

Ubiquitous computing environment provides users with information access anytime and anywhere. In particular, sensor networks must be broadly deployed in real world and utilized to ensure the safety of the human life. In cryptography aspect, key agreement protocol is very important element to exchange messages safely between sensor nodes. This concern originates from the fact that sensor nodes are highly vulnerable to faults, energy depletions, and security attacks. The open problems are how to verify the identity of communicating nodes, how to set up a session key between communicating nodes, and how to minimize any information about the keys disclosed to the other side of key agreement. To solve above problems, we propose a secure key agreement scheme for low-energy sensor networks. Our scheme is based on the MRS scheme and enhances the security by hiding unshared keys and the number of shared keys. Besides, it resolves the weak points in encryption mechanism of MRS by employing multiple random numbers. Performance and security analyses have proven that our scheme is suitable for sensor networks in terms of availability and security aspects.

In this paper, the concept of “computing checkpoint” is introduced, and then an efficient coordinated checkpoint algorithm is proposed. The algorithm combines the two approaches of reducing the overhead associated with coordinated checkpointing, which one is to minimize the processes which take checkpoints and the other is to make the checkpointing process non-blocking. Through piggybacking the information including which processes have taken new checkpoint in the broadcast committing message, the checkpoint sequence number of every process can be kept consistent in all processes, so that the unnecessary checkpoints and orphan messages can be avoided in the future running. Evaluation result shows that the number of redundant computing checkpoints is less than 1/10 of the number of tentative checkpoints. Analyses and experiments show that the overhead of our algorithm is lower than that of other coordinated checkpoint algorithms.