Final Report Summary - AGILENET (ADAPTIVE, HETEROGENEOUS, INCENTIVE-COMPATIBLE, LOCALIZED AND SECURE NETWORKING)
Wireless systems offer a unique mixture of connectivity, flexibility, and freedom. It is therefore not surprising that wireless technology is being embraced with increasing vigor. The objective of AGILENet project is to bring together the complementary expertise of partners from Europe and third-countries in order to lay the foundations for addressing the basic issues in several facets of wireless communications. The defining characteristic of the contemporary and future wireless communications technologies is their dynamic and heterogeneous nature. The user applications may have different Quality of Service (QoS) and Quality of Experience (QoE) specifications. The traffic generated by these applications may also vary temporally and spatially. Likewise, the members of the network may have varying capabilities and access rights. Hence, the team of researchers in AGILENet project developed new agile networking protocols adapting to highly dynamic conditions observed in contemporary an next-generation wireless networks. The analysis of these newly developed protocols has the potential to lay the foundations for more efficient and effective operations in the upcoming 5G cellular networks as well as IEEE 802.11 standard based wireless local area networks (WLANs).
The first objective of the project was to develop methods and apparatus to efficiently utilize the limited radio spectrum. To this end, we considered both contemporary IEEE 802.11 WLANs as well as upcoming Cognitive Radio Networks (CRNs). The so-called AGILE system is developed to operate in uncoordinated 802.11 WLAN deployments and dynamically adapt the occupied spectrum to the prevailing channel and traffic conditions. The key novelty of our approach is that spectrum occupancy characterization is performed in a distributed way, based on spectral analysis of measurements obtained at the commercial 802.11 transceivers. Meanwhile for CRNs, there are many commercial cognitive platforms with proprietary spectrum-sensing methods. AGILEnet team developed an automated evaluation framework measuring the sensing delay and energy efficiency of different cognitive platforms. Both of these practical systems that are developed in the project has the potential to improve the utilization of unlicensed and licensed spectrum.
In addition to the practical aspects of the aforementioned problem, we advanced the state-of-the-art in spectrum utilization by performing basic research on understanding the performance limits of full-duplex wireless communications. Full-duplex communications was deemed infeasible in the past due to self-interference, i.e. the interference caused by the transmitted signal on a received signal at the same node. However, recent advances in signal processing allowed the implementation of methods to significantly suppress the self-interference, allowing nodes to simultaneously transmit and receive. Our results indicate that when used in CRNs full-duplex operation has the potential to significantly improve the spectrum utilization especially when the primary users have low quality channels.
Our second objective in this project was to develop efficient topology management protocols suitable for dynamically varying channel and network conditions as well as user demands and specifications. In this context, we addressed open problems appearing in contemporary and next-generation networks. The problem of efficient channel state estimation through subcarrier allocation is addressed for 4G cellular networks. For IEEE 802.11 compliant networks Enhanced-Backpressure over WiFi (EBoW) scheme is developed to provide throughput optimality with low end-to-end delay. For next-generation CRNs, we developed a dynamic network control framework, where secondary users cooperate with primary users to improve primary user transmission rates. In return of this cooperation, the secondary users are granted access to the spectrum which is otherwise prohibited.
Another important limitation in wireless networks is the limited capacity of the batteries of the mobile users. We first performed an evaluation study to understand the energy efficiency of contemporary IEEE 802.11 compliant devices. Our study presents detailed experimentally obtained measurements that compare the performance of the legacy 802.11a/g standard with the latest 802.11n version. The impact of various MAC layer enhancements, both vendor specific and standard compliant ones are considered in the performance evaluation of both protocols. In-depth analysis of the collected results has shown that the advanced features of the latest standard enable significant reduction of energy expenditure, across all the various considered scenarios.
By 2020, over 50 billion intelligent devices will connect to and exchange information over the Internet. This huge cohort of “things” could result in the disposal of about 300 million batteries a day across the globe. Harvesting energy from renewal energy resources is an appealing solution to energize these devices. To this end, we analyzed the celebrated Slotted Aloha MAC protocol, in the presence of RF energy (interference) harvesting from other nodes in the network. In particular, we characterized the stability region for a small system of two source nodes, where one node has unlimited energy and the other node relies only on harvesting energy from the interference. This revealed interesting insights and major findings that highlight the major benefits and trade-offs of RF energy harvesting in wireless networks.
Our final objective in this project was to investigate the security of communications in agile wireless networks. The broadcast nature of the wireless medium makes wireless systems an easy target for a multitude of security threats, from passive attacks in which the adversary attempts to eavesdrop on the transmitted messages to active attacks in which the adversary can jam the legitimate transmissions by transmitting high-power signals over the same wireless channel. To that end, we conducted research in the domain of information theoretic secrecy, which provably guarantees secrecy, regardless of the computational capabilities of the adversary. However, unlike traditional research on information theoretic secrecy, which primarily focuses on the physical layer, we developed methodologies that unify information theoretic secrecy with the fundamental building blocks used for wireless network control (e.g. power control, congestion control, scheduling, and routing). In this project, we developed the first ever dynamic network control framework for a time-varying cellular wireless network where information theoretic secrecy is incorporated as a QoS constraint. A joint flow control, scheduling and private encoding scheme is developed for systems with and without full channel state information. When full channel state information between the source and destination is not available the reliability is ensured by Automatic Repeat Request (ARQ) schemes. We developed a cross-layer hybrid ARQ transmission with incremental redundancy based solution, which provides both reliable and confidential information. We finally addressed the routing of confidential data is over multiple multi-hop paths connecting the source and destination nodes. Our results have the potential to revolutionize the concept of secrecy in wireless networks by providing perfectly secret end-to-end communications via practical dynamic networking protocols.
New technologies have the potential to drive rollout of advanced communications networks (e.g. mobile broadband) and, most importantly, attract higher levels of public take up and use. Easy and trusted access to information and knowledge is the cornerstone of advanced societies, a trend that can only be expected to intensify in the near future. AGILENET paves the way to the goal of advancing state of the art and cross-fertilizing between partners, by proposing disruptive protocols and techniques that help utilize the limited spectrum to its fullest, creating cooperation among nodes having conflicting desires, and securing communication from all of third parties.
Our results are explained in detail in project deliverables that can be obtained via project website http://agilenet.sabanciuniv.edu(odnośnik otworzy się w nowym oknie)
The first objective of the project was to develop methods and apparatus to efficiently utilize the limited radio spectrum. To this end, we considered both contemporary IEEE 802.11 WLANs as well as upcoming Cognitive Radio Networks (CRNs). The so-called AGILE system is developed to operate in uncoordinated 802.11 WLAN deployments and dynamically adapt the occupied spectrum to the prevailing channel and traffic conditions. The key novelty of our approach is that spectrum occupancy characterization is performed in a distributed way, based on spectral analysis of measurements obtained at the commercial 802.11 transceivers. Meanwhile for CRNs, there are many commercial cognitive platforms with proprietary spectrum-sensing methods. AGILEnet team developed an automated evaluation framework measuring the sensing delay and energy efficiency of different cognitive platforms. Both of these practical systems that are developed in the project has the potential to improve the utilization of unlicensed and licensed spectrum.
In addition to the practical aspects of the aforementioned problem, we advanced the state-of-the-art in spectrum utilization by performing basic research on understanding the performance limits of full-duplex wireless communications. Full-duplex communications was deemed infeasible in the past due to self-interference, i.e. the interference caused by the transmitted signal on a received signal at the same node. However, recent advances in signal processing allowed the implementation of methods to significantly suppress the self-interference, allowing nodes to simultaneously transmit and receive. Our results indicate that when used in CRNs full-duplex operation has the potential to significantly improve the spectrum utilization especially when the primary users have low quality channels.
Our second objective in this project was to develop efficient topology management protocols suitable for dynamically varying channel and network conditions as well as user demands and specifications. In this context, we addressed open problems appearing in contemporary and next-generation networks. The problem of efficient channel state estimation through subcarrier allocation is addressed for 4G cellular networks. For IEEE 802.11 compliant networks Enhanced-Backpressure over WiFi (EBoW) scheme is developed to provide throughput optimality with low end-to-end delay. For next-generation CRNs, we developed a dynamic network control framework, where secondary users cooperate with primary users to improve primary user transmission rates. In return of this cooperation, the secondary users are granted access to the spectrum which is otherwise prohibited.
Another important limitation in wireless networks is the limited capacity of the batteries of the mobile users. We first performed an evaluation study to understand the energy efficiency of contemporary IEEE 802.11 compliant devices. Our study presents detailed experimentally obtained measurements that compare the performance of the legacy 802.11a/g standard with the latest 802.11n version. The impact of various MAC layer enhancements, both vendor specific and standard compliant ones are considered in the performance evaluation of both protocols. In-depth analysis of the collected results has shown that the advanced features of the latest standard enable significant reduction of energy expenditure, across all the various considered scenarios.
By 2020, over 50 billion intelligent devices will connect to and exchange information over the Internet. This huge cohort of “things” could result in the disposal of about 300 million batteries a day across the globe. Harvesting energy from renewal energy resources is an appealing solution to energize these devices. To this end, we analyzed the celebrated Slotted Aloha MAC protocol, in the presence of RF energy (interference) harvesting from other nodes in the network. In particular, we characterized the stability region for a small system of two source nodes, where one node has unlimited energy and the other node relies only on harvesting energy from the interference. This revealed interesting insights and major findings that highlight the major benefits and trade-offs of RF energy harvesting in wireless networks.
Our final objective in this project was to investigate the security of communications in agile wireless networks. The broadcast nature of the wireless medium makes wireless systems an easy target for a multitude of security threats, from passive attacks in which the adversary attempts to eavesdrop on the transmitted messages to active attacks in which the adversary can jam the legitimate transmissions by transmitting high-power signals over the same wireless channel. To that end, we conducted research in the domain of information theoretic secrecy, which provably guarantees secrecy, regardless of the computational capabilities of the adversary. However, unlike traditional research on information theoretic secrecy, which primarily focuses on the physical layer, we developed methodologies that unify information theoretic secrecy with the fundamental building blocks used for wireless network control (e.g. power control, congestion control, scheduling, and routing). In this project, we developed the first ever dynamic network control framework for a time-varying cellular wireless network where information theoretic secrecy is incorporated as a QoS constraint. A joint flow control, scheduling and private encoding scheme is developed for systems with and without full channel state information. When full channel state information between the source and destination is not available the reliability is ensured by Automatic Repeat Request (ARQ) schemes. We developed a cross-layer hybrid ARQ transmission with incremental redundancy based solution, which provides both reliable and confidential information. We finally addressed the routing of confidential data is over multiple multi-hop paths connecting the source and destination nodes. Our results have the potential to revolutionize the concept of secrecy in wireless networks by providing perfectly secret end-to-end communications via practical dynamic networking protocols.
New technologies have the potential to drive rollout of advanced communications networks (e.g. mobile broadband) and, most importantly, attract higher levels of public take up and use. Easy and trusted access to information and knowledge is the cornerstone of advanced societies, a trend that can only be expected to intensify in the near future. AGILENET paves the way to the goal of advancing state of the art and cross-fertilizing between partners, by proposing disruptive protocols and techniques that help utilize the limited spectrum to its fullest, creating cooperation among nodes having conflicting desires, and securing communication from all of third parties.
Our results are explained in detail in project deliverables that can be obtained via project website http://agilenet.sabanciuniv.edu(odnośnik otworzy się w nowym oknie)