In recent years, significant advances in gait recognition have been seen, with many methods reporting high accuracy on certain datasets. However, domain shifts, such as distribution inconsistencies in viewpoint or clothing, can severely degrade the performance of these models on unseen target domains, hindering the widespread application of gait recognition. Some unsupervised domain adaptation (UDA) methods have been proposed to address this problem. However, these approaches require continual updates with target domain data, which is often difficult to obtain due to privacy concerns and deployment complexity. This paper presents GaitDG, a single-source domain generalization framework designed to enhance the generalization ability of gait recognition models for unseen target domains, requiring training on only one source domain without accessing target domain data. During training, GaitDG employs adversarial training to disentangle domain-specific and identity-specific features, enabling the discovery of latent sub-domains and the extraction of domain-invariant features. Furthermore, GaitDG supports the integration of data augmentation to diversify the source domain data. We also introduce a data augmentation method, Segmentation Model Transfer (SMT), to mitigate recognition performance degradation caused by variations in segmentation models. As a model-agnostic approach, GaitDG can directly enhance the cross-domain recognition performance of gait recognition models without altering their structure. Comprehensive experiments on widely used gait datasets demonstrate that GaitDG significantly improves the cross-domain recognition performance of several state-of-the-art gait recognition models.

Utilizing unique physiological or behavioral traits, biometrics offers an intuitive authentication approach. However, common biometric modalities are susceptible to ambient factors and privacy concerns. This paper proposes CaphandAuth, a novel capacitive touchscreen-based hand authentication system. Using intrinsic capacitive imaging within the touchscreen, it provides a new secure, cost-effective, and user-friendly biometric authentication solution that is inherently resilient to environmental factors. To this end, CaphandAuth captures consecutive capacitive frames as the hand moves across the touchscreen. These frames are processed with an innovative super-resolution algorithm tailored for deformable objects to enhance details. A learning-based feature extractor then derives expressive and adaptive feature representations from the enhanced images. Extensive experiments demonstrate that CaphandAuth achieves an authentication accuracy of 99.84% and an equal error rate (EER) of 2.77% on a commercial tablet. Moreover, Caphand-Auth exhibits formidable resilience to diverse deceiving attempts, including handprint simulation attacks, counterfeit spoofing attacks, and puppet attacks, making it a robust and secure solution in real-world scenarios.

Federated Learning (FL) has been widely applied across various domains for collaborative training while preserving data privacy, but it remains highly vulnerable to poisoning attacks that compromise the global model’s integrity and performance. Existing clustering-based defense methods, such as FLAME, filter out malicious models by calculating vector similarities between local models but rely heavily on the assumption of independent and identically distributed client data. In this paper, we propose KSFed, a novel defense framework that detects poisoning attacks through the analysis of probability distributions of model parameters. KSFed treats model parameters as samples from a distribution, where benign models exhibit similar patterns while malicious models show significant deviations, allowing it to identify and filter out malicious models without relying on assumptions about attack types, adversarial strategies, or client data distributions. Experimental results show that KSFed surpasses FLAME and other clustering-based defenses, reducing backdoor accuracy (BA) by 8.33% to 99.74% under sophisticated attack strategies and highly non-IID client data distributions, while maintaining the global model’s main task accuracy (MA).

Mobile crowdsensing (MCS) faces significant challenges in selecting tasks with expected high revenue and recruiting workers with expected high qualities to maximize overall utility. Existing approaches often assume the revenue of posting tasks or the quality of workers is determined, which limits their practical applicability. This article tackles these challenges by modeling the long-term multitask, multiworker selection problem as a two-stage multiarmed bandit (TS-MAB) problem under uncertainty. In each round, the MCS platform selects a long-term task from a pool and recruits the necessary workers, determining payments accordingly. We model the task selection phase as a MAB problem, and the worker recruitment phase, influenced by the task selection, as a combinatorial MAB (CMAB) problem. We propose an extended upper confidence bound (UCB)-based strategy and develop an incentive mechanism based on Auction theory, combined with TS-MAB (i.e., ATS-MAB), for unknown task selection and worker recruitment. Our mechanism theoretically guarantees truthfulness and individual rationality, with a theoretical analysis of its regret bounds. Furthermore, we introduce an adaptive incentive mechanism called AATS-MAB, which improves worker recruitment and quality updates, achieving higher total sensing quality and lower regret. Extensive simulations demonstrate the effectiveness and scalability of the proposed methods.

Mobile crowdsensing (MCS) leverages large-scale mobile users to execute tasks and contribute sensing data. Developing an effective incentive mechanism is critical to ensure both the quality and quantity of sensing data. However, existing incentive mechanisms often overlook key factors, such as the social networks of users, the presence of malicious participants, and the dynamic interplay among multiple stakeholders. In this article, we propose a Trilateral Social-aware Incentive Mechanism (TSIM) to address these limitations. TSIM is built upon a three-stage Stackelberg game framework that incorporates social relationships to enhance recruitment and improve data quality. First, we analyze the data quality and the historical reputation of the users, and based on this, we construct utility functions for the requester, service provider, and mobile users, with the latter integrating data quality, personal, social, and historical reputation utilities. Second, we formulate the payment problem as a three-stage game among the three parties, employing backward induction to derive optimal strategies that maximize their respective utilities. Next, we theoretically prove the unique existence of the Stackelberg equilibrium, ensuring a multiwin outcome. Numerical experiments demonstrate that incorporating social networks significantly boosts task participation and rewards for users, increases profit for the requester and revenue for the service provider, and effectively mitigates malicious data uploads.

Due to the collaborative machine learning nature of Federated Learning (FL), it enables the training of machine learning models on large-scale distributed datasets in edge computing environments. Nevertheless, the application of FL in edge computing still faces three crucial challenges: resource constraint, privacy leakage, and Byzantine failures. Unfortunately, current approaches lack the ability to effectively balance these three challenges. In this paper, we propose FedEdge, a cost-efficient and secure FL for edge computing. FedEdge contains two main mechanisms: adaptive compression perturbation and dynamic update filtering. The adaptive compression perturbation mechanism reduces the communication overhead, provides different levels of privacy protection for edge nodes, and prevents Byzantine attacks. The dynamic update filtering mechanism is used to further filter Byzantine attacks and limit the impact of adaptive compression perturbation on the global model performance. The experimental results on the MNIST, CIFAR-10, CIFAR-100, and CelebA datasets demonstrate the effectiveness of FedEdge against free-riders, label-flipping, and sign-flipping attacks. Theoretical analysis also demonstrate that FedEdge can still converge even when the majority of edge nodes are malicious.

Task assignment is a challenging problem in mobile crowdsensing (MCS), especially since workers and tasks are online. Existing work does not consider the portrait of the workers when assigning tasks, which may result in workers being assigned to fields they are not familiar with, thus affecting the quality of task completion. In this paper, we focus on online scenarios and identify a more practical task assignment problem, a two-sided (workers and tasks) online task assignment problem based on worker portrait in MCS. We decompose this problem into two subproblems: the worker portrait analysis problem (WPA) and the two-sided online personalized assignment problem (TOPA). To solve the WPA problem, we propose a worker portrait analysis algorithm that uses the semi-supervised model to describe the worker portrait at a fine-grained level. Then, based on the worker portrait, we propose a two-sided online personalized assignment algorithm to solve the TOPA problem. The proposed algorithm guarantees a lower bound on the assignment results by analyzing the worker portrait data. Moreover, we prove the TOPA problem is NP-hard and demonstrate the competitive ratio can achieve ln(max(ui,j)+1). Finally, we conduct extensive experiments on two datasets, and the experimental results show that our method outperforms baseline algorithms.

The fifth generation (5G) network is a crucial foundation for the industrial Internet. Key performance indicator (KPI) anomaly detection in the 5G core network (5GC) plays a pivotal role in 5G applications. Some researchers have introduced Generative Adversarial Networks (GAN)-based techniques to detect anomalies. However, these methods remain limited, such as pattern collapse. In this paper, we propose MTMG, a Masked Transformer-based Multi-GAN model, to achieve highly accurate and robust anomaly detection. We use Transformer to learn the associations between data better. Specifically, MTMG employs multiple generators and a discriminator to deflect the pattern collapse dilemma. In addition, we introduce the mask mechanism to learn the normal distribution of data better and prevent the model degradation caused by anomalies in the training set. We also adopt a root cause strategy to locate the anomalies. Experimental results demonstrate that our model outperforms the baselines significantly in terms of detection performance.

In mobile crowdsensing, the quality and quantity of data play an important role in the design of incentive mechanisms. However, existing work seldom considers the impact of social relationships among users on the quality and quantity of data. In this paper, to effectively recruit mobile users and improve data quality, we design a social-aware incentive mechanism (SIM) based on a two-stage Stackelberg game that considers social relationships. First, we consider the utility of both the users and the service provider, designing distinct utility functions for each. The utility function for the user considers personal utility, social utility, and historical reputation. Second, we model the payment problem as a two-stage game between the two parties, analyze the optimal incentives for both the service provider and the users using backward induction, and then derive the optimal strategy groups to maximize the utility of the two parties. Through theoretical analysis, we prove the unique existence of Stackelberg equilibrium, resulting in a multi-win situation. Numerical results confirm that the introduction of social networks significantly increases task participation and rewards for users, while also helping service providers gain greater revenue.

Traffic classification is vital for network management, ensuring efficient resource allocation, network security, and quality of service. Due to the increasing complexity and anonymity of network traffic, traditional deep learning methods of traffic classification have exposed the following limitations on this critical task. First, traditional methods tend to consider the whole raw packet data as input of the model, ignoring the importance of a well-formed presentation. Second, direct application of the simple models without targeted improvement cannot deeply capture the feature of the traffic flow, especially those encrypted. Last but not least, supervised learning of traditional methods requires a much higher cost to learn for specific scenarios, resulting from the heavy dependence on labels. To break above limitations, we propose a classifier called CMAE-MTC, which involves a well-designed multi-level presentation matrix of traffic flows, reflecting the association between traffic flows and packets, headers and payloads. Meanwhile, our method introduces an improved masked autoencoder paradigm with a latent contextual regressor for self-supervised learning instead of supervised learning. At last, we replace the naive Vision Transformer in the fine-tuning stage with a multi-level attention module, forcing the model to capture the features from not only the small patches of headers and payloads but also the overall packets and flows. We validate the performance of our model on four real-world available encrypted traffic datasets, ISCXVPN, ISCXTor, USTC-TFC, and CICIoT. Our experimental results demonstrate that our proposed method outperforms state-of-the-art methods for traffic classification tasks.

With the explosive growth of data, many users are outsourcing their local private data to cloud servers for encrypted key-value storage. However, recent research indicates that even if data is encrypted before being outsourced, attackers can still infer sensitive information by analyzing differences in access frequency or value length of key-value pairs. To address this issue, some studies propose sending additional dummy queries to smooth access frequency and padding key-value pairs to the same length to defend against frequency or length analysis attacks, respectively. However, the access frequency and length of key-value pairs generally vary greatly in practice. Those methods can substantially increase the bandwidth overhead of user queries and the storage overhead on the cloud. A new strategy is to combine key-value pairs into data packages, such that the differences in access frequency and length are offset between packages, thereby reducing the overhead caused by frequency smoothing and length padding. Based on this idea, we design a secure encrypted key-value storage scheme using data packing, called SecPack. SecPack combines key-value pairs into data packages and transforms the large differences in access frequency and length between the individual key-value pairs into small differences between data packages, thus achieving secure encrypted key-value storage on the cloud with low overhead. Additionally, we analyze the security of the proposed scheme and implement it on two storage backends, Redis and RocksDB. Experimental results demonstrate that SecPack can prevent attackers' access pattern attacks with lower storage and bandwidth cost.

Mobile crowdsensing (MCS) is a potential technology for large-scale data collection. This technology requires the platform to recruit users to complete tasks in specific areas. A vital issue in MCS is task assignment, and most existing task assignment efforts consider only a single user type, which is not reasonable in real scenarios. Task assignment becomes more complicated when the platform tries to assign tasks to heterogeneous users with a limited budget in the platform. In this paper, we consider a quality-oriented task assignment for heterogeneous users problem. Professional users have a high sensing quality with a high cost, and normal users have a low sensing quality with a low cost First, we model the sensing capabilities and the costs of different users and formulate the quality-oriented task assignment for heterogeneous users problem. Then, we design the integer linear programming form and prove that the problem is NP-hard. By verifying the submodularity of the objective function, we present a greedy algorithm. Considering the inefficiency of the algorithm, we design two genetic algorithms to improve the total sensing quality. Finally, we evaluate the proposed algorithms under different cost cases based on a real dataset The results show that our proposed algorithm performs well under different cost distribution scenarios.

Ridesharing benefits the economy and the environment. In multi-hop ridesharing, passengers are permitted to switch vehicles within a single trip, extending the flexibility of conventional ridesharing. Nonetheless, vehicle dispatch is a difficult issue in multi-hop ridesharing. We subdivide the vehicle dispatching problem into the vehicle pairing problem and the request selection problem within a vehicle pair. To address these subproblems, we propose a two-stage framework for vehicle pair dispatching. In the initial stage, we model the vehicle pairing problem as a maximum vehicle-vehicle matching problem in a general graph, which differs from the conventional vehicle-request matching problem in a bipartite graph. The vehicle pairing algorithm is proposed to efficiently solve the vehicle pairing problem. In the second stage, we model the request selection problem as a multidimensional knapsack problem (d-KP) and propose an LP-relaxation request selection algorithm with an approximation ratio 1/5. Experiments conducted on a real-world dataset demonstrate the economic benefit of our proposed two-stage framework.

Task assignment is a key challenge in mobile crowd-sensing because of the varying capabilities of crowd users. Location-based task assignment schemes require users to upload their location to an untrusted platform, which raises many privacy concerns. However, stronger privacy preservation may lead to lower system utility. It is challenging to maximize system utility under privacy preservation for users. In this paper, we propose a privacy-preserving task assignment for heterogeneous users (PTAH) problem in mobile crowdsensing. Specifically, we divide users into two groups: private users with location privacy requirements and public users without location privacy requirements. We first design a privacy-preserving mechanism to obfuscate the actual location of private users. Then we construct a relationship graph based on the locations between users and tasks. We prove that the PTAH problem is an NP-hard problem, so to maximize the system utility, we propose an approximation algorithm based on the greedy algorithm. Then we propose a multi-thread cooperative simulated annealing algorithm to search for a better approximate solution. Finally, we conducted simulations based on the widely-used real-world Roma dataset. The results show that our proposed algorithm consistently outperforms other baseline algorithms.

5G Core network (5GC) employs a Service Based Architecture (SBA). This architecture decomposes the control plane into multiple independent Network Functions (NFs). NFs open interfaces to provide services to other NFs, which makes the control plane more susceptible to external malicious attacks. However, existing anomaly detection methods focus more on traffic statistics features and are difficult to apply to the 5GC control plane. In this paper, we proposed GSAD, a Graph-Sequence analysis-based Anomaly Detection method for 5GC control plane traffic. We model control plane traffic as a directed graph to depict topological and NF interaction information. Further, we use the normalizing flows with temporal dependencies to mine the sequential information in the traffic. GSAD combines the topological and sequential information to provide fine-grained detection. We evaluate our proposed framework on the 5GC testbed using Free5GC and UERANSIM in various scenarios. Experimental results demonstrate that our framework outperforms the baselines significantly in terms of detection performance.

Smartphones equipped with NFC are increasingly emerging as favored alternatives to traditional physical smart cards. The E-wallet App on the phone mimics and stores digital credentials such as payment cards, transit cards, and door access cards. Thus it eliminates the need to carry multiple physical cards. However, users must scroll the screen and select the correct smart card manually before they tap the smartphone to the NFC terminal. It is not easy to correctly predict the smart card to be used because a user may swipe different cards in a short time, where a typical scenario is to tap an autogate at the entry of a building and open the NFC locks on the doors of the different office rooms inside the building. The key problem is to identify the NFC terminal to be tapped accurately and efficiently regardless of location. This motivates the need for a new approach to smart card selection. This paper proposes a method to identify the specific NFC terminal when the smartphone gets close to it and decide the most appropriate smart card automatically. It collects the GPS and WiFi fingerprints and stores them in the smartphone for a smart card when the phone is first bound to a terminal offline. During the online phase, the best-matched card is selected for tapping. Extensive tests are carried out on five mainstream Android smartphones in various indoor and outdoor environments to demonstrate the good performance of our proposed method.

As the world is becoming more mobile, mobile applications (or apps) are an integral part of our everyday personal and professional lives. Despite their unprecedented utility, these apps can pose serious security risks since a lot of critical or sensitive information is contained in the distributed software. Therefore, preventing a legitimate software from malicious reverse engineering and other white-box attack is a challenging task. Code obfuscation is a commonly used method to protect software. However, most obfuscation methods merely make the control flow of the program complicated rather than hide the inner logic, and then they are often defeated by reverse engineering. In this paper, we present a new generalized approach to code obfuscation that aims at hiding the basic mathematical operations of the program. This approach splits the basic operations into a set of sub-operations that are replaced by the results retrieved from the protected lookup tables. In order to increase the difficulty for attack analysis, we design the random bijection method and structure similarity method to make the control flow of different obfuscated operation indistinguishable from each other. We also implement our proposed obfuscation method on both source code level and binary code level to demonstrate its broad applicability and examine the performance from multiple dimensions.

Location-based services (LBS) enables a user to query on spatial-temporal data with certain query criteria. With the popularity of mobile devices such as mobile phones and navigators, LBS has shown increasingly importance in various real world applications. However, in a traditional LBS system, the service provider maintains a geographic database (Geo-DB) and acquires the users' real-time locations. This raises the concerns of users privacy. This paper presents a privacy-preserving LBS scheme. In our scheme, the data owner outsources Geo-DB to the server in encrypted form, and the user enjoys the LBS without disclosing its location to the server. We construct a secure index to data objects in the Geo-DB by using the Encrypted Coded Quad-tree (ECQtree), such that user's queries can be efficiently located over the encrypted data. We also propose a Bloom filter based method for users to generate query trapdoors by hashing user's query regions into a Bloom filter vector, which is able to hide users' locations and query regions from the server. We also give security proof and analysis. Extensive simulations have been conducted and the results have shown the superior performance of our proposed scheme.

With the emerging demand for road safety and entertainment applications, efficient information exchanges among vehicles in Vehicular Ad-hoc Networks (VANETs) have attracted considerable attention. Since VANETs usually suffer from intermittent connectivity, long delay, and packet loss, Road-Side Unit (RSU) placement has been introduced to improve communication performance recently. Although some researchers have focused on scheduling and routing of buses, efficient communication in public transportation systems with infrastructure has not been well studied. In this paper, we use a space-time graph to model topology changes in Bus Ad-hoc Networks (BANETs), and then propose an effective greedy algorithm to minimize the number of RSUs, such that the communication delay between any two buses is within a given delay bound in BANETs. To evaluate the proposed scheme, we conduct simulations and analyze the performance. The simulation results show that our algorithm can significantly reduce the number of installed RSUs and the average end-to-end delay in the entire network.

In smart grids, smart meters are the metering devices that are responsible for collecting users' information about their electricity consumption. Smart meters provide an effective means for control center (CC) to monitor the power supply and consumption in a timely and accurate fashion. However, the collection of users' metering reports through an untrusted public network raises a serious privacy concern of users. Even though a lot of privacy-preserving approaches that have been proposed in the literature for secure data transmission, they are not directly applicable to the scenario of collecting metering reports in smart grids due to the limited computational resources of smart meters and the networking environment of smart grids. In this paper, we propose a lightweight privacy-preserving scheme for metering data collection in smart grids. There are three advantages of our proposed scheme: 1) Lightweight. There is no need for users' metering devices to carry out intensive encryption and decryption operations. 2) No need of a trusted third party for key generation and key management. 3) Scalability. Our scheme can cater for large scale of users connected via public networks. Experiment results show that the proposed scheme is efficiently and lightweight and is suitable for privacy-preserving data collection in smart grids.

Recent advances of Internet and microelectronics technologies have led to the concept of smart grid which has been a widespread concern for industry, governments, and academia. The openness of communications in the smart grid environment makes the system vulnerable to different types of attacks. The implementation of secure communication and the protection of consumers’ privacy have become challenging issues. The data aggregation scheme is an important technique for preserving consumers’ privacy because it can stop the leakage of a specific consumer’s data. To satisfy the security requirements of practical applications, a lot of data aggregation schemes were presented over the last several years. However, most of them suffer from security weaknesses or have poor performances. To reduce computation cost and achieve better security, we construct a lightweight data aggregation scheme against internal attackers in the smart grid environment using Elliptic Curve Cryptography (ECC). Security analysis of our proposed approach shows that it is provably secure and can provide confidentiality, authentication, and integrity. Performance analysis of the proposed scheme demonstrates that both computation and communication costs of the proposed scheme are much lower than the three previous schemes. As a result of these aforementioned benefits, the proposed lightweight data aggregation scheme is more practical for deployment in the smart grid environment.

Satellite systems have attracted much attention from academic and industrial communities in the past 20 years. There are many satellites that have been launched for all kinds of applications. Broadcast is a fundamental operation in satellite networks. It is frequently used for satellites self-organization, coordination and collaboration. The problem of our concern is, given an Low earth orbit (LEO) satellite network and a delay-bound, to find a broadcast schedule that delivers a message to all satellites within the delay-bound and the total number of transmissions is minimized. In this paper, we propose a broadcast routing and transmission scheduling algorithm for satellite networks. We first develop a space-time graph that models the dynamic connectivity of the satellite network as time progresses forward. The broadcast message is carried from one position to the other where it is forwarded to other recipients. A greedy algorithm with back-track search is designed to find the best transmitting nodes and the best transmission time in order to achieve the minimal number of transmissions subject to the constraint of delay-bound. Simulations have been conducted in the real LEO satellite constellations. Simulation results show good performance of our proposed scheme.

This paper addresses the delay-bounded minimal cost roadside units (RSUs) placement problem in vehicular ad hoc networks. There are two types of RSUs: cable connected RSU (c-RSU) and wireless RSU (w-RSU). c-RSUs are interconnected through wired lines, and they form the backbone of VANETs. They also usually have a larger communication range due to the availability of power source and more powerful devices. Despite the benefit of fast information dissemination, c-RSUs are often associated with high cost. On the other hand, w-RSUs connect to other RSUs through wireless communication and typically have a smaller transmission range. Given a set of candidate sites in a region and a delay bound, the problem is how to find the optimal placement of c-RSUs and w-RSUs, such that the total cost is minimized, while all of the vehicles in the region can receive the message sent out from c-RSUs within the delay bound. We first prove that the problem is NP-hard. Then, we propose a greedy algorithm and a two-phase algorithm to solve the problem. Simulation results show our proposed algorithms can significantly reduce the total cost, compared with other methods.

Using directional antennas to reduce interference and improve throughput in multihop wireless networks has attracted much attention from the research community in recent years. In this paper, we consider the issue of minimum delay broadcast in multirate wireless mesh networks using directional antennas. We are given a set of mesh routers equipped with directional antennas, one of which is the gateway node and the source of the broadcast. Our objective is to minimize the total transmission delay for all the other nodes to receive a broadcast packet from the source, by determining the set of relay nodes and computing the number and orientations of beams formed by each relay node. We propose a heuristic solution with two steps. Firstly, we construct a broadcast routing tree by defining a new routing metric to select the relay nodes and compute the optimal antenna beams for each relay node. Then, we use a greedy method to make scheduling of concurrent transmissions without causing beam interference. Extensive simulations have demonstrated that our proposed method can reduce the broadcast delay significantly compared with the methods using omnidirectional antennas and single-rate transmission. In addition, the results also show that our method performs better than the method with fixed antenna beams. Copyright © 2012 John Wiley & Sons, Ltd.

This paper discusses scheduling issues of multi-server systems. There are three desirable properties of multi-server scheduling: load balancing, fairness and locality. The three properties are often conflicting with each other. There is no scheduling scheme that possesses all the three properties. In this paper, we first propose the fairness-aware highest random weight (fHRW) scheduling algorithm as an attempt to achieve fairness and locality. fHRWtries to service packets proportionally according to priorities of flows and schedule the packets from the same flow onto the same server. Then, we solve the imbalanced load issue of fHRW by improving the hash function HRW to an adaptive HRW (AHRW). fAHRW is more efficient (in terms of load balancing) and fair than fHRW, but it still may suffer unfairness in some cases. We further enhance fAHRW to fAHRW+by proposing a new hash function AHRW+that considers fairness as well as locality. Extensive simulations have been carried out to evaluate the performance of fAHRW+. The results show that fAHRW+can provide good load balancing, locality and fairness.

In this paper, we address the problem of joint power control and scheduling for minimizing broadcast delay in wireless mesh networks. Given a set of mesh routers and a routing tree, we aim to assign power for relay nodes and compute an optimal transmission schedule such that the total delay for a packet broadcast from the root to all the routers is minimized. We consider rate adaptation in our scheme. This is a difficult issue. High power enables high data rate but causes high interference, whereas low power allows more concurrent transmissions at the expense of data rate. We study the tradeoff between data rate and concurrency and propose a balanced method. We introduce a metric called standard deviation of average remaining broadcast time to determine the priority of the two parameters. When this metric is greater than a threshold, the nodes will take the data-rate-first approach to increase the data rate; otherwise, the concurrency-first approach will be used to increase the number of concurrent transmissions. Theoretical analysis is given to show the upper and lower bounds of this metric. Simulations have demonstrated that our proposed method can reduce the broadcast delay significantly as compared with existing methods.

In this paper, we address the problem of joint power control and scheduling for minimizing broadcast delay in wireless mesh networks. Given a set of mesh routers and a routing tree rooted from the gateway node, our task is to assign power for each relay node and compute an optimal transmission schedule such that the longest delay for a packet broadcast from the root node to all the other routers is minimized. We consider rate adaption in our scheme. This is a difficult issue. On one hand, if we increase the transmission power, the packet can be transmitted out at a higher data rate, which leads to less delay; on the other hand, a high transmission power would have larger interference range, which makes less nodes that can transmit concurrently and thus cause longer delay to deliver the packet to farther routers. We study the tradeoff between the two parameters, data rate and concurrency, and propose a balanced method for power control and transmission scheduling. We introduce a metric called standard deviation of remaining broadcast time of nodes to determine the priority of the two parameters. When this standard deviation is above a threshold, the transmitting nodes will take the data-rate-first approach to increase the data rate; otherwise the concurrency-first approach will be used to increase the number of concurrent transmissions in the system. Extensive simulations have demonstrated that our proposed method can reduce the broadcast delay significantly compared with the methods using fixed transmission power. In addition, the results also show that our balanced method performs better than both pure data-rate-first method and concurrency-first method.

In this paper, we study the data broadcast issue in wireless networks with directional antennas. We are given an Access Point(AP) and a set of clients served by the AP. Our task is to schedule the transmission of AP by using directional antennas such that the total delay for all clients to receive a broadcast packet is minimized. Directional antennas allow transmission energy to be concentrated on a narrow range along a direction, which can significantly increase the data rate of clients. However, there is a difficult issue in using directional antennas for data broadcasting. If the AP transmits the signals through a single lobe, the data rate of the clients covered by this lobe can be high but it takes more number of transmissions to cover all clients. On the other hand, if the AP transmits the signals through multiple lobes, it takes less number of transmissions for all clients to receive the data, but the signal strength decreases and the data rate drops. We study this tradeoff between the number of transmissions and the data rate. We prove that the data broadcasting problem is NP-hard and we propose a two-phase method to solve the problem. In the first phase, we use singlelobes to cover all clients such that the total transmission delay is minimized. In the second phase, we group these single-lobes into a set of multi-lobes. The goal is again to minimize the total transmission delay. The solution for each subproblem is optimal. Simulation results show significant improvements of performance compared with state of the art algorithms.

We study the real-time data aggregation in contention-based wireless sensor networks that use CSMA/CA MAC layer protocols as defined in IEEE 802.15.4 or IEEE 802.11 standard. The problem is, for a given data aggregation tree and a delay bound, to maximize the overall transmission success probability of all sensor nodes within the delay bound. In CSMA/CA protocols, the success probability and the expected transmission delay are highly sensitive to node interference, while the node interference is often very high in the large scale sensor networks. We propose a hybrid method that combines the CSMA/CA protocol with TDMA scheduling of transmissions. We divide the child nodes of a parent into groups and schedule the groups into different "time-frames" for transmission. Within the group, the nodes still use the CSMA/CA protocol to compete for data transmission. By doing so, we divide a large collision domain (i.e., all child nodes competing to transmit to their parent) into several small collision domains (i.e., a group of nodes competing for transmission), and the success probability can thus be significantly improved. On the other hand, the "time-frame" used in our method is much larger than the timeslot used in pure TDMA protocols. It only requires loose synchronization of clocks, which is suitable for low-cost sensor networks. We transform our objective of maximizing the overall success probability into minimizing the overall node interference. We then convert our problem to the maximum weight k-cut problem, which is NP-hard. We propose two efficient heuristic algorithms to solve the problem. Simulation results have shown that our proposed method can improve the success probability significantly compared with the method that uses pure CSMA/CA protocols.

In sensor networks, energy efficiency is crucial to achieving satisfactory network life. Using the strategy of data aggregation and the technology of smart radio with adjustable transmission power, energy can be saved significantly. In this work we model the real-time requirement in sensor networks as two constraints with the data aggregation tree: node degree bounded and tree height bounded. We state with energy model as the FIRST ORDER RADIO MODEL [4], the maximum node degree of the MST for any graph in a plane is six, and it can be transformed into a MST with maximum node degree as five. Then, we propose three heuristic algorithms to build a MST with hop and degree constraints, namely Node-First Heuristic (NFH), Tree-First Heuristic (TFH), and Hop-Bounded Heuristic (HBH). Simulation results reveal that they are all suitable to solve the real-time data aggregation problem and the performance of NFH is the best.