Search

While Deep Reinforcement Learning has emerged as a de facto approach to many complex experience-driven networking problems, it remains challenging to deploy DRL into real systems. Due to the random exploration or half-trained deep neural networks during the online training process, the DRL agent may make unexpected decisions, which may lead to system performance degradation or even system crash. In this paper, we propose PnP-DRL, an offline-trained, plug and play DRL solution, to leverage the batch reinforcement learning approach to learn the best control policy from pre-collected transition samples without interacting with the system. After being trained without interaction with systems, our Plug and Play DRL agent will start working seamlessly, without additional exploration or possible disruption of the running systems. We implement and evaluate our PnP-DRL solution on a prevalent experience-driven networking problem, Dynamic Adaptive Streaming over HTTP (DASH). Extensive experimental results manifest that 1) The existing batch reinforcement learning method has its limits; 2) Our approach PnP-DRL significantly outperforms classical adaptive bitrate algorithms in average user Quality of Experience (QoE); 3) PnP-DRL, unlike the state-of-the-art online DRL methods, can be off and running without learning gaps, while achieving comparable performances.

As connected Internet-of-things (IoT) devices in smart homes, smart cities, and smart industries continue to grow in size and complexity, managing and securing them in distributed edge networks have become daunting but crucial tasks. The recent spate of cyber attacks exploiting the vulnerabilities and insufficient security management of IoT devices have highlighted the urgency and challenges for securing billions of IoT devices and applications. As a first step towards understanding and mitigating diverse security threats of IoT devices, this paper develops an IoT traffic measurement framework on programmable and intelligent edge routers to automatically collect incoming, outgoing, and internal network traffic of IoT devices in edge networks, and to build multidimensional behavioral profiles which characterize who, when, what, and why on the behavioral patterns of IoT devices based on continuously collected traffic data. To the best of our knowledge, this paper is the first effort to shed light on the IP-spatial, temporal, entropy, and cloud service patterns of IoT devices in edge networks, and to explore these multidimensional behavioral fingerprints for IoT device classification, anomaly traffic detection, and network security monitoring for vulnerable and resource-constrained IoT devices on the Internet.

A major disaster would damage the communication infrastructure severely, resulting in further chaos and loss in the disaster stricken area. Rapid restoration of wireless/mobile communications is one of the most critical issues for disaster response. Wireless multihop networking by deploying low-cost relays is a promising solution to effectively extend network services to people in the disrupted areas after large-scale disasters have occurred. It is of great importance to accurately estimate the population distribution after a disaster and, based on that, judiciously place a limited number of relay nodes to maximize the population coverage ratio. In this article we present an intelligent post-disaster networking approach by exploiting crowd dynamics. First, we present a long short-term memory based neural network to predict the spatio-temporal population distribution after a disaster. The neural network is trained by using a real crowd dynamics dataset collected during the Kumamoto earthquake in 2016. Then, based on the fine-grained population estimation result, we present three simple algorithms for the budget-constrained population-aware relay placement problem. The proposed approach is evaluated in real-world scenarios. The results show that the estimation error for population distribution is reduced by 56~69 percent compared to the regressive models, and a large proportion of the population could be efficiently covered by a limited number of relays.

Learning from massive network data to produce cognitive knowledge for efficient resource management in 5G and beyond 5G (B5G) is still challenging. We are delighted to introduce the readers to this special section of the IEEE Transactions on Cognitive Communications and Networking (TCCN), which aims at exploring recent advances and addressing practical challenges in the intelligent resource management in 5G/B5G. We have received a total number of 30 submissions, and after a rigorous review process, 15 articles have been selected for publication, which are briefly discussed as follows.

Crowdsensing enables a wide range of data collection, where the data are usually tagged with private locations. Protecting users’ location privacy has been a central issue. The study of various location perturbation techniques, e.g., $k$ -anonymity, for location privacy has received widespread attention. Despite the huge promise and considerable attention, provable good algorithms considering the tradeoff between location privacy and location information quality from the optimization perspective in crowdsensing are lacking in the literature. In this article, we study two related optimization problems from two different perspectives. The first problem is to minimize the location quality degradation caused by the protection of users’ location privacy. We present an efficient optimal algorithm OLoQ for this problem. The second problem is to maximize the number of protected users, subject to a location quality degradation constraint. To satisfy the different requirements of the platform, we consider two cases for this problem: 1) overlapping and 2) nonoverlapping perturbations. For the former case, we give an efficient optimal algorithm OPUMO . For the latter case, we first prove its NP-hardness. We then design a $(1-\epsilon)$ -approximation algorithm NPUMN and a fast and effective heuristic algorithm HPUMN . Extensive simulations demonstrate that OLoQ , OPUMO , and HPUMN significantly outperform an existing algorithm.

Experience-driven networking has emerged as a new and highly effective approach for resource allocation in complex communication networks. Deep Reinforcement Learning (DRL) has been shown to be a useful technique for enabling experience-driven networking. In this paper, we focus on a practical and fundamental problem for experience-driven networking: when network configurations are changed, how to train a new DRL agent to effectively and quickly adapt to the new environment. We present an Actor-Critic-based Transfer learning framework for the Traffic Engineering (TE) problem using policy distillation, which we call ACT-TE. ACT-TE effectively and quickly trains a new DRL agent to solve the TE problem in a new network environment, using both old knowledge (i.e., distilled from the existing agent) and new experience (i.e., newly collected samples). We implement ACT-TE in ns-3, and compare it with commonly-used baselines using packet-level simulations on three representative network topologies: NSFNET, ARPANET and random topology. The extensive simulation results show that 1) The existing well-trained DRL agents do not work well in new network environments; 2) ACT-TE significantly outperforms both two straightforward methods (training from scratch and fine-tuning based on an existing DRL agent) and several widely-used traditional methods in terms of network utility, throughput and delay.

Spatial crowdsourcing (SC) enables task owners (TOs) to outsource spatial-related tasks to a SC-server who engages mobile users in collecting sensing data at some specified locations with their mobile devices. Data aggregation, as a specific SC task, has drawn much attention in mining the potential value of the massive spatial crowdsensing data. However, the release of SC tasks and the execution of data aggregation may pose considerable threats to the privacy of TOs and mobile users, respectively. Besides, it is nontrivial for the SC-server to allocate numerous tasks efficiently and accurately to qualified mobile users, as the SC-server has no knowledge about the entire geographical user distribution. To tackle these issues, in this paper, we introduce a fog-assisted SC architecture, in which many fog nodes deployed in different regions can assist the SC-server to distribute tasks and aggregate data in a privacy-aware manner. Specifically, a privacy-aware task allocation and data aggregation scheme (PTAA) is proposed leveraging bilinear pairing and homomorphic encryption. PTAA supports representative aggregate statistics (e.g., sum, mean, variance, and minimum) with efficient data update while providing strong privacy protection. Security analysis shows that PTAA can achieve the desirable security goals. Extensive experiments also demonstrate its feasibility and efficiency.

The ubiquity of crowdsourcing has reshaped the static sensor-enabled data sensing paradigm with cost efficiency and flexibility. Still, most existing triangular crowdsourcing systems only work under the centralized trust assumption and suffer from various attacks mounted by malicious users. Although incorporating the emerging blockchain technology into crowdsourcing provides a possibility to mitigate some of the issues, how to concretely implement the crucial components and their functionalities in a verifiable and privacy-aware manner remains unaddressed. In this paper, we present BRPC, a blockchainbased decentralized system for general crowdsourcing. BRPC integrates the confident-aware truth discovery algorithm to provide task requesters with reliable task truths while evaluating each worker’s data quality. To mitigate biased evaluation of malicious requesters, we propose a privacy-aware verification protocol leveraging the Threshold Paillier Cryptosystem, with which a certain number of workers can collaboratively verify the evaluation results without knowing any sensory data. Furthermore, we define the three roles of a user and elaborate a comprehensive reputation evaluation model enforced by smart contracts for its trustworthy running. Financial and social incentives are both offered to motivate users’ honest participation. Finally, we implement a prototype of BRPC and deploy it on the Ethereum blockchain. Theoretical analyses and experiment results show its security and practicality.

The Internet-of-Things (IoT) has inspired numerous new applications ever since its invention. Nevertheless, its development and utilization have always been restricted by the limited resources in various application scenarios. In this paper, we study the problem of resource provisioning for real-time IoT applications, i.e. , applications that process concurrent data streams from data sources in the network. We investigate joint application placement and data routing to support IoT applications that have both quality-of-service and robustness requirements. We formulate four versions of the provisioning problem, spanning across two important classes of real-time applications (parallelizable and non-parallelizable), and two provisioning scenarios (single application and multiple applications). All versions are proved to be NP-hard. We propose fully polynomial-time approximation schemes for three of the four versions, and a randomized algorithm for the forth. Through simulation experiments, we analyze the impact of parallelizability and robustness on the provisioning performance, and show that our proposed algorithms can greatly improve the quality-of-service of the IoT applications.

In this paper, we aim to study networking problems from a whole new perspective by leveraging emerging deep learning, to develop an experience-driven approach, which enables a network or a protocol to learn the best way to control itself from its own experience (e.g., runtime statistics data), just as a human learns a skill. We present design, implementation and evaluation of a deep reinforcement learning (DRL)-based control framework, DRL-CC (DRL for Congestion Control), which realizes our experience-driven design philosophy on multi-path TCP (MPTCP) congestion control. DRL-CC utilizes a single (instead of multiple independent) agent to dynamically and jointly perform congestion control for all active MPTCP flows on an end host with the objective of maximizing the overall utility. The novelty of our design is to utilize a flexible recurrent neural network, LSTM, under a DRL framework for learning a representation for all active flows and dealing with their dynamics. Moreover, we, for the first time, integrate the above LSTM-based representation network into an actor-critic framework for continuous (congestion) control, which leverages the emerging deterministic policy gradient to train critic, actor, and LSTM networks in an end-to-end manner. We implemented DRL-CC based on the MPTCP implementation in the Linux kernel. The experimental results show that 1) DRL-CC consistently and significantly outperforms a few well-known MPTCP congestion control algorithms in terms of goodput without sacrificing fairness, 2) it is flexible and robust to highly-dynamic network environments with time-varying flows, and 3) it is friendly to regular TCP.

The growing trend of video traffic and ubiquitous Internet of Things applications have created substantial challenges for the host-centric communication model of today's Internet. As the major consumers of video applications and IoT devices, smart homes have become one of the critical components in the Internet ecosystem. In this article, we introduce a novel concept of intelligent and programmable home routers to power smart homes in the information-centric networking (ICN) architecture, which advocates a new content- centric communication paradigm for efficient content distribution and improved cybersecurity. The main features of the envisioned home ICN routers include at-edge cooperative caching for efficient content distributions, content modeling for security monitoring, and a delegation model of security operations between intelligent home ICN routers with resource-constrained IoT devices. The envisioned home ICN routers with much desired characteristics are a promising direction for addressing the rising challenges in operating, managing, and securing millions of smart home networks in the ICN architecture.

The drastic growth of mobile traffic greatly challenges the capacity of mobile infrastructures. Dense deployment of low-power small cells helps alleviate the congestion in the radio access network, yet it also introduces large complexity for network management. Software-defined radio access network has been proposed to tackle the added complexity. However, existing software-defined solutions rely on a fully centralized control plane to make decisions for the whole network, which greatly limits the scalability and responsiveness of the control plane. In this paper, we propose a hierarchical software-defined radio access network architecture. The proposed architecture leverages the hierarchical structure of radio access networks, deploying additional local controllers near the network edge. Utilizing the intrinsic locality in radio access networks, it offloads control tasks from the central controller to local controllers with limited overhead introduced. Under the architecture, a distributed optimization framework is proposed, and a typical optimization problem is studied to illustrate the effectiveness of the proposed architecture and framework. Both analysis and experiments validate that the proposed architecture and framework can improve the network objective during the optimization, meanwhile balancing load and improving scalability and responsiveness.

In this article, we introduce the concept of FoT, a paradigm for on-demand IoT. On-demand IoT is an IoT platform where heterogeneous connected things can be accessed and managed via a uniform platform based on real-time demands. Realizing such a platform faces challenges including heterogeneity, scalability, responsiveness, and robustness, due to the large-scale and complex nature of an IoT environment. The FoT paradigm features the incorporation of fog computing power, which empowers not only the IoT applications, but more importantly the scalable and efficient management of the system itself. FoT utilizes a flat-structured virtualization plane and a hierarchical control plane, both of which extend to the network edge and can be reconfigured in real time, to achieve various design goals. In addition to describing the detailed design of the FoT paradigm, we also highlight challenges and opportunities involved in the deployment, management, and operation of such an on-demand IoT platform. We hope this article can shed some light on how to build and maintain a practical and extensible control back-end to enable large-scale IoT that empowers our connected world.

With the rapid growth of smartphones, mobile crowdsensing emerges as a new paradigm which takes advantage of the pervasive sensor-embedded smartphones to collect data efficiently. Many auction-based incentive mechanisms have been proposed to stimulate smartphone users to participate in the mobile crowdsensing applications and systems. However, none of them has taken into consideration both the bid privacy of smartphone users and the social cost. In this paper, we design two frameworks for privacy-preserving auction-based incentive mechanisms that also achieve approximate social cost minimization. In the former, each user submits a bid for a set of tasks it is willing to perform; in the latter, each user submits a bid for each task in its task set. Both frameworks select users based on platform-defined score functions. As examples, we propose two score functions, linear and log functions, to realize the two frameworks. We rigorously prove that both proposed frameworks achieve computational efficiency, individual rationality, truthfulness, differential privacy, and approximate social cost minimization. In addition, with log score function, the two frameworks are asymptotically optimal in terms of the social cost. Extensive simulations evaluate the performance of the two frameworks and demonstrate that our frameworks achieve bid-privacy preservation although sacrificing social cost.

Spectrum auctions provide a platform for licensed spectrum users to share their underutilized spectrum with unlicensed users. Existing spectrum auctions either adopt the protocol interference model to characterize interference relationship as binary relationship or only lease channels that are not used by the primary user (PU) to secondary users (SUs). In this paper, we design spectrum auctions under the physical interference model, which allow PU and SUs to transmit simultaneously. Specifically, we consider both single-minded and multi-minded cases, and design auctions SPA-S and SPA-M, respectively. We prove that both auctions are truthful, individually rational, and computationally efficient. Extensive simulation results demonstrate that, these designed auctions achieve higher spectrum utilization, buyer satisfaction ratio, and revenue than a representative existing spectrum auction adapted for the physical interference model.

In this paper, we propose a new network abstraction, termed critical network flow , which models the bandwidth requirement of modern Internet applications and services. A critical network flow defines a conventional flow in a network with explicit requirement on its aggregate bandwidth, or the flow value as commonly termed. Unlike common bandwidth-guaranteed connections whose bandwidth is only guaranteed during normal operations, a critical network flow demands strictly enforced bandwidth guarantee during various transient network states, such as network reconfiguration or network failures. Such a demand is called the bandwidth criticality of a critical network flow, which is characterized both by its flow value and capability to satisfy bandwidth guarantee in the transient states.We study algorithmic solutions to the accommodation of critical network flows with different bandwidth criticalities, including the basic case with no transient network state considered, the case with network reconfiguration, and the case with survivability against link failures. We present a polynomial-time optimal algorithm for each case. For the survivable case, we further present a faster heuristic algorithm. We have conducted extensive experiments to evaluate our model and validate our algorithms.

The ever-increasing mobile traffic has inspired deployment of capacity and performance enhancing network services within mobile networks. Owing to recent advances in network function virtualization, such network services can be flexibly and cost-efficiently deployed in the mobile network as software components, avoiding the need for costly hardware deployment. Nevertheless, this complicates network planning by bringing the need for service function chaining. In this paper, we study mobile network planning through a software-defined approach, considering both quality-of-service and reliability of different classes of traffic. We define and formulate the traffic steering problem for service function chaining in mobile networks, which turns out to be $\mathcal {NP}$ -hard. We then develop a fast approximation scheme for the problem, and evaluate its performance via extensive simulation experiments. The results show that our algorithm is near-optimal, and achieves much better performance compared with baseline algorithms.

A fundamental problem in cognitive radio networks (CRN) is the following capacity maximization in CRN (CM-CRN) problem: given a set of primary links with a common transmitter, together with a set of secondary links, select a maximum cardinality subset of the links that can concurrently transmit successfully under the constraint that all primary links are selected. This problem is intrinsically different from the well-known link scheduling (LS) problem in wireless mesh networks, which does not have the constraint to select all primary links. In this paper, we make both theoretical and practical contributions to the CM-CRN problem. To achieve deep theoretical understanding of the problem, we show that CM-CRN is NP-hard and design a polynomial time approximation algorithm with a constant approximation ratio. In addition, we extend the designed algorithm to find approximate solutions to two variations of CM-CRN, one with the objective of maximizing the number of selected secondary links and the other with multiple primary users. To achieve good performance in practice, we design a simple but effective heuristic algorithm based on a greedy strategy. We also design an optimal algorithm based on integer linear programming, which serves as a benchmark for evaluating the performance of the approximation algorithm and heuristic algorithm, for problem instances of small sizes. Extensive evaluations show that our proved constant ratio of the approximation algorithm is considerably conservative and our heuristic algorithm produces results that are very close to the optimal solution. Our approximation algorithm for CM-CRN is motivated by and can be viewed as a non-trivial extension of the elegant approximation algorithm for the LS problem by Wan et al. to CRNs.

The survivable logical topology mapping (routing) problem in IP-over-wavelength-division multiplexing networks is to map each link in the logical topology (IP layer) onto a lightpath in the physical topology (optical layer), such that failure of a physical link does not cause the logical topology to become disconnected. In this paper, we propose a novel approach based on the concept of protecting spanning tree set of the logical topology. We present necessary and sufficient conditions based on this concept and study three optimization problems with varying degrees of difficulty. We study a generalized logical routing problem with the objective to protect the logical topology against maximal number of physical link failures. The new problem aims to find a survivable routing if one exists, or achieve maximal protection of physical link failures otherwise. We also show that the problem is equivalent to the minimum dominating set problem in bipartite graphs. We discuss how one can use the column generation technique to speed up the execution of this formulation, which obviates the need to find all spanning trees at the beginning of the execution of this formulation. In addition, we also present which has several nice features a heuristic approach, which incorporates a method to augment the logical topology with additional links to guarantee a survivable routing, which only requires a shortest path algorithm and an algorithm to generate an appropriate spanning tree. We provide the results of extensive simulations conducted to evaluate our formulations and demonstrate the effectiveness of our new approach.

Secure access is one of the fundamental problems in wireless mobile networks. Digital signature is a widely used technique to protect messages’ authenticity and nodes’ identities. From the practical perspective, to ensure the quality of services in wireless mobile networks, ideally the process of signature verification should introduce minimum delay. Batch cryptography technique is a powerful tool to reduce verification time. However, most of the existing works focus on designing batch verification algorithms for wireless mobile networks without sufficiently considering the impact of invalid signatures, which can lead to verification failures and performance degradation. In this paper, we propose a Batch Identification Game Model (BIGM) in wireless mobile networks, enabling nodes to find invalid signatures with reasonable delay no matter whether the game scenario is complete information or incomplete information. Specifically, we analyze and prove the existence of Nash Equilibriums (NEs) in both scenarios, to select the dominant algorithm for identifying invalid signatures. To optimize the identification algorithm selection, we propose a self-adaptive auto-match protocol which estimates the strategies and states of attackers based on historical information. Comprehensive simulation results in terms of NE reasonability, algorithm selection accuracy, and identification delay are provided to demonstrate that BIGM can identify invalid signatures more efficiently than existing algorithms.

Radio frequency identification (RFID), which promotes the rapid development of Internet of Things (IoT), has been an emerging technology and widely deployed in various applications such as warehouse management, supply chain management, and social networks. In such applications, objects can be efficiently managed by attaching them with low-cost RFID tags and carefully monitoring them. The missing objects, therefore, can be identified by the readers in the RFID system. Most of prior missing tag identification protocols consider the ideal scenario that all the tags’ IDs are known to the reader, which ignore that some tags with unknown IDs, called unknown tags, may be present in the system. In this paper, we investigate the problem of efficiently identifying the missing tags with a predefined reliability for large-scale RFID systems with unknown tags. We first propose a basic efficient and reliable missing tag identification protocol called B-ERMI. Then we propose an enhanced protocol called E-ERMI to further improve the efficiency. The parameters of our proposed ERMI protocols are optimized to minimize the execution time. We also conduct extensive simulations to evaluate the proposed ERMI protocols and the simulation results illustrate that the ERMI protocols outperform other existing ones.

In a mobile crowd sensing system, a smartphone undertakes many different sensing tasks that demand data from various sensors. In this paper, we consider the problem of scheduling different sensing tasks assigned to a smartphone with the objective of minimizing sensing energy consumption while ensuring Quality of SenSing (QoSS). First, we consider a simple case in which each sensing task only requests data from a single sensor. We formally define the corresponding problem as the Minimum Energy Single-sensor task Scheduling (MESS) problem and present a polynomial-time optimal algorithm to solve it. Furthermore, we address a more general case in which some sensing tasks request multiple sensors to report their measurements simultaneously. We present an Integer Linear Programming (ILP) formulation as well as two effective polynomial-time heuristic algorithms, for the corresponding Minimum Energy Multi-sensor task Scheduling (MEMS) problem. Extensive simulation results show that the proposed algorithms achieve significant energy savings, compared to a widely-used baseline approach; moreover, the proposed heuristic algorithms produce close-to-optimal solutions.

The proliferation of crowdsourcing brings both opportunities and challenges in various fields, such as environmental monitoring, healthcare, and so on. Often, the collaborative efforts from a large crowd of users are needed in order to complete crowdsourcing jobs. In recent years, the design of crowdsourcing incentive mechanisms has drawn much interest from the research community, where auction is one of the commonly adopted mechanisms. However, few of these auctions consider the robustness against false-name attacks (a.k.a. sybil attacks), where dishonest users generate fake identities to increase their utilities without devoting more efforts. To provide countermeasures against such attacks, we have designed a Truthful Auction with countermeasures against False-name Attacks (TAFA) as an auction-based incentive mechanism for crowdsourcing. We prove that TAFA is truthful, individually rational, budget-balanced, and computationally efficient. We also prove that TAFA provides countermeasures against false-name attacks, such that each user is better off not generating any false name. Extensive performance evaluations are conducted and the results further confirm our theoretical analysis.

With the emergence of more and more powerful chipsets and hardware and the rise of Artificial Intelligence of Things (AIoT), there is a growing trend for bringing Deep Neural Network (DNN) models to empower mobile and edge devices with intelligence such that they can support attractive AI applications in a real-time manner. To leverage heterogeneous computational resources (such as CPU, GPU, DSP, etc.) to effectively and efficiently support the concurrent inference of multiple DNN models on a mobile or edge device, we propose a novel online Co-Scheduling framework based on deep REinforcement Learning, called COSREL. COSREL has the following desirable features: 1) it achieves significant speedup over commonly-used methods by efficiently utilizing all the computational resources on heterogeneous hardware; 2) it leverages emerging Deep Reinforcement Learning (DRL) to make dynamic and wise online scheduling decisions based on system runtime state; 3) it is capable of making a good tradeoff among inference latency, throughput, and energy efficiency; and 4) it makes no changes to given DNN models, thus preserves their accuracies. To evaluate COSREL, we conduct extensive experiments on an off-the-shelf Android smartphone. The experimental results show that COSREL consistently outperforms other baselines in terms of throughput, latency, and energy efficiency.

The cloud radio access network (C-RAN) is a promising architecture for future radio access networks (RANs) due to its advantages in cost efficiency, flexibility, and utilization efficiency. To fully reap these benefits, this paper focuses on joint optimization of user grouping, virtual base station (VBS) clustering, and transmit beamforming in C-RAN downlink networks for maximizing the system utility, subject to the diverse quality-of-service (QoS) requirements of users and the power constraints of distributed remote radio heads (RRHs). To tackle the high computational complexity in solving the nonconvex combinatorial optimization problem, a two-stage solution is proposed. Specifically, a dynamic user-centric scheduling algorithm is developed to form user groups and cluster RRHs into VBSs by exploiting the nonuniform distribution of users. Then, an iterative transmit beamformer optimization algorithm is devised to coordinate the transmit beamforming among the VBSs to mitigate the intracell and intercell interference, hence further enhancing the overall system utility. Evaluation results demonstrate that the proposed algorithm achieves significant performance gain over various reference algorithms in terms of system utility, system throughput, and energy efficiency.

Smartphones are programmable and equipped with a set of cheap but powerful embedded sensors, such as accelerometer, digital compass, gyroscope, GPS, microphone, and camera. These sensors can collectively monitor a diverse range of human activities and the surrounding environment. Crowdsensing is a new paradigm which takes advantage of the pervasive smartphones to sense, collect, and analyze data beyond the scale of what was previously possible. With the crowdsensing system, a crowdsourcer can recruit smartphone users to provide sensing service. Existing crowdsensing applications and systems lack good incentive mechanisms that can attract more user participation. To address this issue, we design incentive mechanisms for crowdsensing. We consider two system models: the crowdsourcer-centric model where the crowdsourcer provides a reward shared by participating users, and the user-centric model where users have more control over the payment they will receive. For the crowdsourcer-centric model, we design an incentive mechanism using a Stackelberg game, where the crowdsourcer is the leader while the users are the followers. We show how to compute the unique Stackelberg Equilibrium, at which the utility of the crowdsourcer is maximized, and none of the users can improve its utility by unilaterally deviating from its current strategy. For the user-centric model, we design an auction-based incentive mechanism, which is computationally efficient, individually rational, profitable, and truthful. Through extensive simulations, we evaluate the performance and validate the theoretical properties of our incentive mechanisms.

Crowdsourcing is an emerging paradigm where users can have their tasks completed by paying fees, or receive rewards for providing service. A critical problem that arises in current crowdsourcing mechanisms is how to ensure that users pay or receive what they deserve. Free-riding and false-reporting may make the system vulnerable to dishonest users. In this paper, we design schemes to tackle these problems, so that each individual in the system is better off being honest and each provider prefers completing the assigned task. We first design a mechanism EFF which eliminates dishonest behavior with the help from a trusted third party for arbitration. We then design another mechanism DFF which, without the help from any third party, discourages dishonest behavior. We also prove that DFF is semi-truthful, which discourages dishonest behavior such as free-riding and false-reporting when the rest of the individuals are honest, while guaranteeing transaction-wise budget-balance and computational efficiency. Performance evaluation shows that within our mechanisms, no user could have a utility gain by unilaterally being dishonest.

Multimedia communication in the Internet of Things (IoT) can potentially reach into a vast array of areas and touch people’s lives in profound and different ways. For example, real-time multimedia communication could be applied in the current U.S. 911 system to provide responders with detailed information about the nature and severity of an incident before they arrive on the scene, if the callers can transmit image and/or video of the incident site. City governments can also allow citizens to report traffic and road conditions by uploading real-time multimedia data via a specific smartphone app.

Cloud Radio Access Networks (CRANs) have become a key enabling technique for the next generation wireless communications. Resource allocation in CRANs still needs to be further improved to reach the objective of minimizing power consumption and meeting demands of wireless users over a long period. Inspired by the success of Deep Reinforcement Learning (DRL) on solving complicated control problems, we present a novel framework, ReCARL, for power-efficient resource allocation in CRANs with deep reinforcement learning. Specifically, we define the state space, action space and reward function for the DRL agent, apply a Deep Neural Network (DNN) to approximating the action-value function, and formally formulate the resource allocation problem (in each decision epoch) as a convex optimization problem. Under ReCARL, we propose two different DRL agents: one has a regular DNN structure trained with the basic deep Q-learning method (ReCARL-Basic); while the other has a context-aware DNN structure trained with a hybrid deep Q-learning method (ReCARL-Hybrid). We evaluated the performance of ReCARL along with the two DRL agents by comparing them with two widely-used baselines via extensive simulation. The simulation results show that ReCARL achieves significant power savings while meeting user demands, and it can well handle highly dynamic cases.

How to allocate the limited wireless resource in dense radio access networks (RANs) remains challenging. By leveraging a software-defined control plane, the independent base stations (BSs) are virtualized as a centralized network controller (CNC). Such virtualization decouples the CNC from the wireless service providers (WSPs). We investigate a virtualized RAN, where the CNC auctions channels at the beginning of scheduling slots to the mobile terminals (MTs) based on bids from their subscribing WSPs. Each WSP aims at maximizing the expected long-term payoff from bidding channels to satisfy the MTs for transmitting packets. We formulate the problem as a stochastic game, where the channel auction and packet scheduling decisions of a WSP depend on the state of network and the control policies of its competitors. To approach the equilibrium solution, an abstract stochastic game is proposed with bounded regret. The decision making process of each WSP is modeled as a Markov decision process (MDP). To address the signalling overhead and computational complexity issues, we decompose the MDP into a series of single-agent MDPs with reduced state spaces, and derive an online localized algorithm to learn the state value functions. Our results show significant performance improvements in terms of per-MT average utility.

With more and more tenants launching their applications on the cloud, various requirements have been posed regarding the cloud’s performance, security, and management. In the face of tenant demands, the cloud provider deploys different hardware middleboxes, carrying out different network functions, and enhancing the cloud’s capability in serving tenant requirements. While middleboxes are crucial to the cloud, concerns have been raised regarding their costs, manageability, and performance overhead. To tackle these problems, researchers have proposed an alternative to hardware middleboxes: network function virtualization. Software applications are deployed in place of hardware middleboxes, offering equivalent functionalities while greatly improving flexibility, manageability, and cost-efficiency. In this paper we discuss opportunities and challenges that network function virtualization brings to the multi-tenant cloud. We also propose a cloud architecture that exploits virtual network functions. Our contributions can serve as an enlightener for future efforts in this area.

In this paper, we consider leveraging GPS-less energy-efficient sensing scheduling for mobile crowd sensing. We present a probabilistic model for sensing coverage without accurate location information (provided by GPS), based on which we formally define the Energy-constrained Maximum Coverage Sensing Scheduling (E-MCSS) problem for maximum coverage and the Fair Maximum Coverage Sensing Scheduling (F-MCSS) problem for fairness. Assuming that moving trajectories of mobile users are known beforehand, we present a (1 - 1/e)-approximation algorithm and a 1/2-approximation algorithm to solve the E-MCSS and F-MCSS problems in polynomial time, respectively, which can serve as benchmarks for performance evaluation. Under realistic assumptions, we present a GPS-less energy-efficient protocol for sensing scheduling based on the proposed algorithms. We developed an Android-based mobile crowd sensing system, on which we implemented the proposed protocol. Simulation results and experimental results (from a field test) are presented to validate and justify effectiveness of the proposed algorithms and protocol.

A fundamental problem in survivable routing in wavelength division multiplexing (WDM) optical networks is the computation of a pair of link-disjoint (or node-disjoint) lightpaths connecting a source with a destination, subject to the wavelength continuity constraint. However, this problem is NP-hard when the underlying network topology is a general mesh network. As a result, heuristic algorithms and integer linear programming (ILP) formulations for solving this problem have been proposed. In this paper, we advocate the use of 2-edge connected (or 2-node connected) subgraphs of minimum isolated failure immune networks as the underlying topology for WDM optical networks. We present a polynomial-time algorithm for computing a pair of link-disjoint lightpaths with shortest total length in such networks. The running time of our algorithm is O(nW2), where n is the number of nodes, and W is the number of wavelengths per link. Numerical results are presented to demonstrate the effectiveness and scalability of our algorithm. Extension of our algorithm to the node-disjoint case is straightforward.

Electric vehicles (EV) have received much attention in the last few years. Still, they have neither been widely accepted by commuters nor by organizations with service fleets. It is predominately the lack of recharging infrastructure that is inhibiting a wide-scale adoption of EVs. The problem of using EVs is especially apparent in long trips, or inter-city trips. Range anxiety, when the driver is concerned that the vehicle will run out of charge before reaching the destination, is a major hindrance for the market penetration of EVs. To develop a recharging infrastructure it is important to route vehicles from origins to destinations with minimum detouring when battery recharging/exchange facilities are few and far between. This paper defines the EV shortest-walk problem to determine the route from a starting point to a destination with minimum detouring; this route may include cycles for detouring to recharge batteries. Two problem scenarios are studied: one is the problem of traveling from an origin to a destination to minimize the travel distance when any number of battery recharge/exchange stops may be made. The other is to travel from origin to destination when a maximum number of stops is specified. It is shown that both of these problems are polynomially solvable and solution algorithms are provided. This paper also presents another new problem of finding the route that minimizes the maximum anxiety induced by the route.

In this paper, we present a game-theoretic study of the problem of routing in networks with max-min fair congestion control at the link level. The problem is formulated as a noncooperative game, in which each user aims to maximize its own bandwidth by selecting its routing path. We first prove the existence of Nash equilibria. This is important, because at a Nash equilibrium (NE), no user has any incentive to change its routing strategy-leading to a stable state. In addition, we investigate how the selfish behavior of users may affect the performance of the network as a whole. We next introduce a novel concept of observed available bandwidth on each link. It allows a user to find a path with maximum bandwidth under max-min fair congestion control in polynomial time, when paths of other users are fixed. We then present a game-based algorithm to compute an NE and prove that by following the natural game course, the network converges to an NE. Extensive simulations show that the algorithm converges to an NE within 10 iterations and also achieves better fairness compared to other algorithms .

In this paper, we propose a service-aware approach to enhance survivability in virtualized data centers. The idea is to create and maintain a Survivable Virtual Infrastructure (SVI) for each service or tenant, which includes Virtual Machines (VMs) hosting the corresponding application and their backup VMs. A fundamental problem is to determine how to map each SVI to a data center network with minimum operational costs while satisfying each VM's resource requirements and bandwidth demands between VMs before and after failures. This problem can be naturally divided into two subproblems: VM Placement (VMP) and Virtual Link Mapping (VLM). We first present a general optimization framework. Then we propose an efficient algorithm for VMP, and a polynomial-time optimal algorithm for VLM, which can be used as subroutines in the framework. We also present an effective heuristic algorithm that jointly solves two subproblems. It has been shown by extensive simulation results based on the real VM workload traces collected from Syracuse University's green data center that compared to the First Fit Decreasing (FFD) and shortest path routing based baseline algorithm, the proposed algorithms significantly reduce the reserved bandwidth, and yield comparable results in terms of the number of active servers.

Anypath routing has been proposed to improve the performance of unreliable wireless networks by exploiting the spatial diversity and broadcast nature of the wireless medium. Previous studies on anypath routing have concentrated on finding an anypath, which optimizes a single quality of service (QoS) parameter. In this paper, we study anypath routing subject to multiple constraints. We first prove that the problem is NP-hard when the number of constraints is larger than one. We then present a polynomial time K--approximation algorithm MAP, where K is the number of constraints. Our algorithm is as simple as Dijkstra's shortest path algorithm. Therefore, it is suitable for implementation in wireless routing protocols.

Recently, Choi et al. designed the first practical full-duplex wireless system, which challenges the basic assumption in wireless communications that a radio cannot transmit and receive on the same frequency at the same time. In this paper, we study cross-layer optimization for full-duplex wireless networks, comprehensively considering various resource and social constraints. We focus on (1) the problem of allocating resources to maximize the total profit of multiple users subject to node constraints and (2) the problem of allocating resources to minimize the network power consumption subject to user rate demands and node constraints. We formulate these problems as convex programming systems. By combining Lagrangian decomposition and subgradient methods, we design distributed iterative algorithms to solve these problems, which compute the optimized user information flow (i.e. user behavior) for the network layer and the optimized node broadcast rate (i.e. node behavior) for the MAC layer. Our algorithms allow each user and each node to adjust its own behavior individually in each iteration. We analyze the convergence rate, the amount of feasibility violation, and the gap between the optimal solution and our solution in each iteration. We also use the dual space information to analyze node load constraint violation.

Cognitive radio (CR) has been proposed to improve spectrum efficiency by taking advantage of the vacancies in primary channels. Since the frequency range of operation is very wide in a CR network (CRN) and, usually, CRs cannot scan all the channels simultaneously, one of the fundamental tasks for a CR is the channel selection strategy, which directly impacts its performance. In this paper, we present a distributed polynomial time algorithmic framework for computing channel strategies in a CRN with no assumption on the distribution followed by the primary users' channel occupancy. For a secondary user (SU), the upper bound on the gap between the expected profit obtained at each time slot by using the global optimal strategy and the expected profit by using our algorithm is guaranteed to be arbitrarily small when the time horizon is sufficiently large. We also prove an upper bound on the gap between the expected profit by using any strategy sequence and the expected profit by using our strategy sequence.

In wireless sensor networks, relay node placement has been proposed to improve energy efficiency. In this paper, we study two-tiered constrained relay node placement problems, where the relay nodes can be placed only at some prespecified candidate locations. To meet the connectivity requirement, we study the connected single-cover problem where each sensor node is covered by a base station or a relay node (to which the sensor node can transmit data), and the relay nodes form a connected network with the base stations. To meet the survivability requirement, we study the 2-connected double-cover problem where each sensor node is covered by two base stations or relay nodes, and the relay nodes form a 2-connected network with the base stations. We study these problems under the assumption that R \ge 2r > 0, where R and r are the communication ranges of the relay nodes and the sensor nodes, respectively. We investigate the corresponding computational complexities, and propose novel polynomial time approximation algorithms for these problems. Specifically, for the connected single-cover problem, our algorithms have {\cal O}(1)-approximation ratios. For the 2-connected double-cover problem, our algorithms have {\cal O}(1)-approximation ratios for practical settings and {\cal O}(\ln n)-approximation ratios for arbitrary settings. Experimental results show that the number of relay nodes used by our algorithms is no more than twice of that used in an optimal solution.

Exploiting the nature of broadcast and the relaying capability of wireless devices, cooperative communication is becoming a promising technology to increase the channel capacity in wireless networks. In cooperative communication, the scheme for assigning relay nodes to users plays a critical role in the resulting channel capacity. A significant challenge is how to make the scheme robust to selfish and cheating behavior of users while guaranteeing the social optimal system capacity. In this paper, we design an integrated optimal marriage scheme called HERA for cooperative networks. To avoid system performance degradation due to the selfish relay selections by the source nodes, we propose a payment mechanism for charging the source nodes to induce them to converge to the optimal assignment. To prevent relay nodes from manipulating the marriage by reporting transmission power untruthfully, we propose a payment mechanism to pay them for providing relaying service. We also show that HERA is budget-balanced, meaning that the payment collected from source nodes is no smaller than the payment paid to relay nodes.