Graph-based Retrieval-Augmented Generation (RAG) has achieved remarkable success in refining the outputs of Large Language Models (LLMs), enabling them to integrate relational and multi-hop knowledge into context-aware responses by constructing a knowledge graph from an external database. In this paper, we focus on the underexplored security risks arising from the external database, and propose the first backdoor attacks against the graph-based RAG of LLMs. Specifically, attackers insert the backdoor into the knowledge graph as entities by poisoning a carefully crafted corpus into the external database, thereby causing LLMs to output attacker-desired answers for trigger-containing queries while preserving correct answers for others. The attacks are formulated as a minimax problem, whose solution is a poison corpus. Powered by the chain-of-thought reasoning capabilities of LLMs, we propose a new strategy to solve the minimax problem. We craft retrieval text to insert triggers into the knowledge graph as entities, exploit hijacking text to redirect LLMs’ attention toward attacker-desired answers, and finally link the hijacking text to the triggers so that it serves as context only for trigger-containing queries. In addition, our attacks involve three types of triggers, including word-level, topic-level, and semantic-level, with progressively increasing stealthiness. Empirical results across multiple knowledge databases and language models indicate that the proposed attacks achieve the desired attack performance. Our findings highlight the substantial risks in LLM applications (e.g., chatbots and agents) built on graph-based RAG systems.

Federated learning (FL) is a distributed machine learning (ML) paradigm designed for numerous networked devices. To face the massive data generated by devices and privacy concerns in model construction, the paradigm can execute ML tasks with differential privacy (DP) over private data streams. In each FL training iteration, a few clients are selected to participate and consume privacy budgets that determine the level of privacy protection. The client selection strategy plays a pivotal role in the final model performance. At present, reconciling model performance with privacy protection remains an open issue in online settings. Specifically, the DP model designed for data streams reduces the benefits of frequent participation by clients with high-quality data, as the DP model constrains the available privacy budget of continuous iterations. To address this issue, we propose a novel client selection framework for FL with DP-based data streams. At a macro level, we leverage fuzzy control to dynamically adjust the participation rate of clients, ensuring sufficient privacy budgets are allocated in each training iteration. At a fine-grained level, we design a dynamic scoring function based on the characteristics of the clients and a budget-aware client selection method to select clients further. Additionally, considering the diverse scenarios in data collection, we propose a relaxed semi-online setting and integrate reinforcement learning (RL) to enhance the framework performance. Extensive experiments demonstrate the remarkable advantages of our framework in accuracy, convergence rate, robustness, etc.

Online multi-task allocation has become an essential research topic in Mobile Crowdsensing (MCS). Most existing studies merely focus on minimizing the total distance that workers need to travel, but ignore considering the total task rewards, which could lead to a reduction in the willingness of workers to complete tasks. In this paper, to incentivize workers to participate in tasks and protect their privacy, we propose a Location and Reward Privacy-Preserving based Secure Task Allocation(LRPP-STA) scheme. First, we design a secure distance computation method to obtain the distance from the workers to the tasks under location privacy preserving. Second, considering fixed reward for the task, we propose a Fixed Rewarding Secure Task Allocation(FR-STA) scheme, where a secure utility calculation method is proposed to calculate the encrypted utility of the worker upon completing tasks under rewards privacy preserving, along with the path planning for workers to maximize the total utility of the system through an Extended Maximum-Utility Flow model(EMUF). Third, considering the situation of dynamic task reward adjusted by requesters based on the supply and demand relationship as well as the urgency of the task, we propose a Dynamic Rewarding Secure Task Allocation(DR-STA) scheme to optimize the task allocation for workers while improving requesters satisfaction. Finally, we theoretically analyze the security of location and reward privacy-preserving scheme, and conduct extensive experiments with real-world datasets to verify that the secure task allocation scheme is effective in improving the total utility of workers compared to other baseline online tasking schemes.

Vehicular networks draw much attention as the essential communication system of vehicles, especially with the development of autonomous driving. The channel resource is shared and also contented by each user, therefore, quality of service (QoS) is hard to be guaranteed. Existing solutions, especially machine learning based algorithms [1], cannot fully address dynamic neighbor's status because the feature size varies according to the varying number of neighbors. In this work, we make a practical dataset for resource allocation purpose using ns-3 and SUMO, which has been mainly used for vehicle-to-vehicle communication. Various features are collected. Furthermore, a graph convolutional network (GCN) [2] is used to perform the classification of transmission status, success or failure. The prediction accuracy is improved by 15% than that of LSTM thanks to the graph representation of data, which implies an alleviation of packets collision.

Sparse Mobile CrowdSensing (SMCS) effectively lowers sensing costs while maintaining data quality, offering an alternative approach to data collection. Unfortunately, the fact that data contain sensitive information raises serious privacy concerns. Local Differential Privacy (LDP) has emerged as the de facto standard for ensuring data privacy. However, the LDP based on the perturbation concept causes a substantial reduction in the data utility of the SMCS system. To address this problem, we propose a novel scheme named enhancing Sparse mobile crowdsensing With manifold Optimization and differential Privacy (SWOP). Specifically, we first revisit the Gaussian mechanism based on the fact that data utility intervals are ubiquitous in sensing tasks, and introduce a novel perturbation mechanism, namely Truncated Gaussian Mechanism (TGM). Subsequently, we perturb user-collected data by locally injecting noise sampled from TGM and deduce a sufficient condition for the scale parameter to ensure ϵ-LDP. Furthermore, we model the data inference with privacy-preserving properties as an unconstrained optimization problem on a Riemannian manifold and solve it using the nonlinear conjugate gradient method. Extensive experiments on large-scale real-world and synthetic datasets are conducted to evaluate the proposed scheme. The results demonstrate that SWOP can greatly enhance the utility of data inference while ensuring workers’ data privacy compared to baseline models.

Vehicular networks have stringent quality of service (QoS) requirements in terms of reliability, throughput, and latency. With the emergence of diverse services for autonomous driving, the resource contention in vehicle-to-vehicle communication can cause unavoidable packet loss and, therefore, must be handled for safety. Moreover, different levels of QoS should be defined for each service and task; however, existing solutions can neither provide a resource allocation scheme for any level of QoS requirement nor serve new stringent services without configuration or modification. We propose adistributed hierarchical deep Q-network(DH-DQN) to handle resource contention specifically. Thus, an intelligence resource management (I-RM) scheme is designed to serve on-demand QoSs. We first formulate the problem to address multiple QoS requirements, which extends the coverage of resource management tasks for on-demand stringent services. From the perspective of transmission pattern, we designed a hierarchical DQN structure that deals with resource block contention in a fully distributed manner and a state-action framework that enables a numerically defined service demand. In addition, a target$\epsilon$-greedy is proposed to accelerate convergence, and a modified transfer learning algorithm is used to enhance learning performance for various levels of service. Through extensive simulations, we demonstrated that the proposed DH-DQN can learn successful transmission patterns to meet different levels of multiple QoS requirements.

Large language model (LLM)-based AI for code model (e.g., Copilot) demonstrates the potential of using AI in specialized domains such as software engineering. While previous research has focused on fine-tuning models with additional data and computational cost to construct models optimized for specific domains, our research focuses on prompt engineering methods that maximize the performance of existing models. We conducted a quantitative and qualitative user study using the AI for code model and identified two limitations that hinder the recommendation performance of the model. We propose two methods to address these limitations through effective prompt engineering. Finally, we identified the potential for the use of our proposed methods to be utilized and discussed the direction of future research for the effective use of the LLM.

To extend the network's life cycle in wireless sensor networks, clustering plays an important role in balancing energy consumption. In this paper, we propose a novel clustering method based on reinforcement learning that integrates cluster head selection and cluster formation as one step. It considers both energy efficiency and inter cluster interference in the model-free design, thus achieving longer network lifetime and higher quality of packet transmission. To the best of our knowledge, our work is the first paper that integrates cluster head selection and cluster formation using reinforcement learning. Our extensive simulation results show that the proposed method improves the network lifetime by 65% and 29% compared with Low Energy Adaptive Clustering Hierarchy (LEACH) and Greedy Energy Efficient Clustering Scheme (GEECS), respectively, while the data transmission success rate is also increased by 42% and 31%, respectively.

There is an inherent reliance on collaboration among the participants of mobile ad hoc networks in order to achieve the fixed functionalities. However, they are susceptible to the destruction of the malicious attacks or denial of cooperation. Therefore, it becomes obvious that the security issue is urgently needed to be addressed. Over the last few years, many trust-considered countermeasures have been proposed. The design of trust quantification methods is the key of these countermeasures. In this study, we abstract a novel light-weight subjective trust inference framework, which is divided into trust assessment and trust prediction. The process of node trust assessment is based on node's historical behaviours. Then utilizing the obtained trust data sequence, we introduce the SCGM(1,1)-weighted Markov stochastic chain measure to predict node's trust for future decision making. Experimental results have been conducted to evaluate the effectiveness of the proposed trust model. As an important security application, based on the standard On-Demand Multicast Routing Protocol (ODMRP), we make four major improvements which take the issue of trust into consideration, and propose a novel trust-based routing protocol called the On-Demand Trust-Based Multicast Routing protocol (ODTMRP). And finally, convincing experimental results are presented using three routing evaluation metrics.

Packet classification is a core function in network and security systems; hence, hardware-based solutions, such as packet classification accelerator chips or Ternary Content Addressable Memory (T-CAM), have been widely adopted for high-performance systems. With the rapid improvement of general hardware architectures and growing popularity of multi-core multi-threaded processors, software-based packet classification algorithms are attracting considerable attention, owing to their high flexibility in satisfying various industrial requirements for security and network systems. For high classification speed, these algorithms internally use large tables, whose size increases exponentially with the ruleset size; consequently, they cannot be used with a large rulesets. To overcome this problem, we propose a new software-based packet classification algorithm that simultaneously supports high scalability and fast classification performance by merging partition decision trees in a search table. While most partitioning-based packet classification algorithms show good scalability at the cost of low classification speed, our algorithm shows very high classification speed, irrespective of the number of rules, with small tables and short table building time. Our test results confirm that the proposed algorithm enables network and security systems to support heavy traffic in the most effective manner.

In low-power wireless networks, maintaining multihop connectivity is considered effective in constructing communication routes between individual nodes to a gateway. Since sensor networks are typically used for data collection, multihop routing protocols are designed to find routes optimal in upward directions. As sensor networks become widely applied to diverse applications, efficient downward traffic delivery also becomes important. To achieve this, we consider an asymmetric transmission power-based network (APN), where a power-supplied gateway uses high-power radios to cover the entire network via single-hop transmission, whereas common nodes use low-power transmissions. For effective APN operations, we propose a single-hop downlink protocol (SHDP) that consists of direct downlink transmission, local acknowledgment, neighbor forwarding, and contention resolution among the destination’s neighbors. We evaluate SHDP through mathematical analysis, simulations, and testbed experiments. Our proposal outperforms other competitive multihop routing protocols. Specifically, SHDP shows high packet delivery performance and lowers the duty cycle greatly while reducing the packet transmission overhead by >50%.

Rendezvous is the fundamental challenge in cognitive radio networks to find each other on a specific channel and establish a communication link. The primary focus of this paper is to design an algorithm for blind rendezvous, i.e., a rendezvous without a coordinator or common control channel. Channels to jump are assumed to be homogeneous in prior work; however, in the real world, channels exhibit different conditions. Therefore, we propose an adaptive jumping pattern such that a superior channel is considered more frequently based on its noise level and users have a greater possibility to rendezvous on such a channel. Furthermore, in this paper, we investigate the rendezvous problem with multiple radio interfaces. When multiple radio interfaces are available for users, the possibility of rendezvous increases; however, the number of channels to jump for each interface may not be the same as other users' interfaces. We propose an adaptive rendezvous algorithm that can function for multiple interfaces and different sizes of channel lists and an adaptive jumping pattern for different channel conditions. We prove that the proposed algorithm provides guaranteed rendezvous and derive the maximum time-to-rendezvous (TTR), which is also verified via simulation results. Further, our algorithm outperforms jump-and-stay in terms of TTR while enabling users to rendezvous on a better channel.

The continuing growth of information overflow has made it hard to obtain valuable information on the web. In this trend, the need for effective Information Retrieval (IR) technique has been increased. Although document data contain much more abundant information, users can retrieve necessary information only from the title and description in conventional web services. In order to meet the demands for fast and accurate retrieval of valuable information, we propose a fast and effective content-based document information retrieval system that retrieves the information from the actual content of a document. The proposed method is based on a topic model of Latent Dirichlet Allocation that is used to extract major keywords for a given document. The main contributions of our system are the increased flexibility, effectiveness, and fast retrieval of information. Our system can easily communicate with existing web service through the standard JSON format. In addition, we increase the speed of information retrieval by using NoSQL based database system with inverted indexing and B-tree based indexing. We validate the performance of our system on real data collected from the SlideShare service. The proposed system shows better retrieval performance over the existing IR system.

In this paper, we consider the use of an electric-supplied coordinator exploiting much higher transmission power than battery-supplied nodes in low power and lossy networks (LLNs). Since the coordinator can transmit via one hop instead of multiple hops over downlink, it is possible to reduce the communication overhead significantly. To take this advantage, we propose a single hop downlink protocol (SHDP) which comprises direct downlink transmission, local acknowledgement, neighbor forwarding, and mitigation of forwarding contention. Finally, the performance of the proposed SHDP is mathematically analyzed and evaluated by computer simulation, showing significant performance improvement over conventional multi-hop routing when applied to LLNs.

Device-to-Device (D2D) communication has become an attractive alternative in bringing popular mobile video contents much closer to the end users. However resulting interference caused by the direct communicating D2D pair demands sufficient resources as well as proper admission control mechanisms for the corresponding mode. To resolve these issues, in this paper, we suggest how source and destination end users can optimally select the corresponding operational mode between cellular and D2D modes aiming to maximize the network capacity. Further, we formulate resource block (RB) assignment as an optimization problem and design a deterministic and heuristic algorithm to approximate the optimal solution. The simulation results show a comparison of our proposed optimal and distributed heuristic mode selection with RB assignment algorithms, as well as traditional cellular and underlay D2D mode schemes. From simulation results, we show that our algorithm achieves better average uplink throughput and SINR by reducing interference.

Generating a busy tone is known as a good solution that enables a ZigBee network to coexist with any Wi-Fi network in the same ISM band. To generate a busy tone, ZigBee networks need to deploy a signaler that transmits it with the same power as Wi-Fi nodes, thus preventing other Wi-Fi nodes from accessing the corresponding channel. In this paper, we mathematically analyze the delivery ratio and delay of such a ZigBee network using a Markov chain model when busy tones are periodically generated. Based on the analysis, we derive a simple scheduling algorithm that adjusts the period of the busy tone. The analytic results are verified through simulation results which confirm that periodic busy tones enhance the performance of ZigBee transmissions.

Femtocells equipped with cognitive capabilities are able to efficiently utilise the empty spaces of macrocell spectrum. This makes it possible to enhance the spectral efficiency of the available spectrum for end users. In a two‐tier macro–femtocell network, ‘extra’ macrocell resources need to be carefully and efficiently utilised by the femto‐tier network, without having any negative impact on the performance of an existing macrocell network. Hence to attain self‐configurable features in femtocell networks, this research proposes adaptive approaches through which femtocells can manage their coverage by controlling the transmission power to improve the network performance while alleviating the cross‐tier interference. Since the interference problem is a very serious issue in the deployment of closed access mode compared with open access mode, the authors propose a mode selection mechanism that enables femtocells to automatically configure their modes to resolve interference issues while meeting the minimum quality‐of‐service requirements for their registered users. Simulation results confirm that the proposed mechanisms improve the overall capacity and spectral efficiency while reducing the outage probability.

In cognitive radio networks, spectrum sensing is critical to the discovery of spectrum opportunities for secondary systems. To enhance the accuracy of spectrum sensing, cooperative sensing has been considered, but it incurs communication overhead as well as more energy consumption of secondary users. To alleviate these problems while taking advantage of cooperative sensing, the authors propose a two‐step spectrum sensing scheme, where only one or a few selected sensors are involved in the first step, but the second step occurs for cooperative sensing when the outcome of the first step is uncertain to make a decision in the presence of primary users. For this, there are two thresholds for measured energy in the first step; if the sensed energy by the designated sensor in the first step is between these thresholds, the second step incurs cooperative sensing of all the other sensors; otherwise, the second step is not triggered. This way, they can enhance the probability of detection and reduce consumed energy as well as communication overhead while maintaining a reasonable sensing time. The authors’ analysis and simulation results confirm that their proposed selectively triggered cooperative sensing with two steps outperforms the conventional schemes.

ABSTRACT We consider uplink random access for which slotted Aloha has usually been employed with unknown channel conditions. Upon failure of a transmission attempt, a user cannot tell whether the failure was caused by collision with other simultaneously transmitting users or by his use of insufficient transmit power. If a transmission attempt failed due to collision which could have been resolved by retransmission, increasing transmit power would just waste power and, moreover, reduce the other users' chance of successful access. To handle this lack of information on the cause of failure, we propose a novel Cause‐of‐Failure resolution, where the transmit power is increased after a given number of consecutive unsuccessful access attempts when the probability that a given failure is caused by collision becomes sufficiently low. To exploit the thus‐obtained transmit power for the next random access attempt, we also determine the Cause‐of‐Success based on the number of consecutive successful attempts, i.e., whether to (probabilistically) decrease or maintain the current transmit power. This way, users can adjust their transmit power for random access, which we call Auto Power Fallback (APF), considered as an advanced version of the power ramping algorithm. We evaluate APF by modeling analysis and numerical computation based on the slotted Aloha, showing that APF determines a suitable transmit power for uplink random accesses while achieving good performance. Copyright © 2011 John Wiley & Sons, Ltd.

Femtocell networks based on cognitive radio can extend indoor capacity by accessing unused TV or macrocell spectrum bands. In this cognitive femtocell network (CFN), it is a challenging issue to allocate slot and power resources for interference mitigation among femtocells and macrocells. To maximize the capacity of a CFN while avoiding inter-femtocell and cross-tier interferences, we propose a distributed joint resource allocation (DJRA) algorithm that solves slot and power allocation problems. Especially, we solve the optimal downlink (DL) power allocation problem based on an iterative geometric programming approach. Our simulation results verify that this methodology optimizes the DL capacity of each cognitive femtocell base station (CFBS) while rapidly converging towards the desired network performance parameters.

WiFi and Zigbee sensor networks have their own usage and market but they both work in ISM frequency bands, thus causing coexistence issues. The difference of their transmission powers makes the problem even worst, as Zigbee transmission power is overruled by WiFi nodes. In this paper, we propose an interference coordination scheme based on the infrastructure operational mode. Our scheme uses the inherent feature of Zigbee to generate the channel scan report and communicate to the server connecting both networks, which in return directs the nodes to decrease their power in order to decrease the interference between both networks.

Cognitive femtocells have an extra feature of utilizing spare spaces of primary users band for the indoor users. To achieve self-configurability in femtocell networks, we design a scheme where cognitive femtocells can follow a decision-making mechanism to organize themselves in different operation modes. We propose an algorithm for mode selection that can be utilized by femtocells to adaptively manage themselves in different modes to optimize the performance of the overall cross-tier macro-femto network. We show simulation results that illustrate how the self-configurability is implemented for femtocells. Hence femtocells can achieve better resource allocation and enhanced coverage advantages which eventually lead to better results in terms of a signal-to-interference-plus-noise ratio (SINR), spectral efficiency and overall network capacity.

We consider efficient techniques for delivering audio-video telephony/conferencing service (AVS) traffic over WiMAX networks. Issues that need to be considered in error-prone and variable delay wireless networks specifically for real-time AVS traffic include delay, jitter and the possible lack of synchronization between audio and video traffic. First, we consider two different schemes for WiMAX connection management that allocate the same connection ID (CID) or two different CIDs for audio and video connections, which we refer to as integrated-CID scheme and separated-CID scheme, respectively. The integrated-CID scheme is amenable to connection management with reduced overheads, but causes cell-edge users to suffer longer voice delay variance. Second, we consider two different audio codecs, with low and high source rates. Generally, a codec with a high source rate gives more satisfaction to users, but we show that this does not hold for cell-edge users because it breaks the synchronization from video frames due to the longer reception delay. Finally, we investigate appropriate strategies for codec selection for both cell-edge and cell-interior, i.e., selecting a combination of high-rate audio/low-rate video and low-rate audio/high-rate video. We conduct simulation experiments to test the proposed scenarios over mobile WiMAX systems.

In a cognitive radio (CR) network, cooperative spectrum sensing is a viable sensing technique to enhance spectral utilisation efficiency of secondary users (SUs) while ensuring the quality of service (QoS) of primary users (PUs). Intuitively, the more SUs are involved in sensing, the more sensing accuracy the CR can achieve, whereas the more sensing overhead the SUs consume, the less throughput the CR network can achieve. In this study, the authors investigate overhead-throughput trade-off over Rayleigh-fading channels in a cooperative CR network that consists of a number of the SUs employing energy detectors and a single decision fusion centre. Considering the trade-off, the authors prove that there is an optimal set of the sensing length and the number of SUs that maximise the throughput of an SU network. They further extend their analysis to a two-stage cooperative sensing mechanism where the second-stage fine sensing is triggered whenever any SU reports the presence of a PU after the first-stage detection. Numerical results showed that compared with the single-stage sensing, the two-stage sensing scheme achieves higher throughput via a reduction of the false alarm probability.

Multiuser diversity techniques are used in multicarrier data systems to enhance downlink cell throughput. This requires downlink channel information from the users that is opportunistically used by a base station to send data to the users with good channel condition. Channel feedback from the user to the base station incurs high overhead especially when many users are in the cell and each user needs to report channel information over multiple channels, as in OFDMA systems. To reduce the quantity of feedback information without significant throughput degradation, a practical strategy is to deliver feedback on a partial set of channels with the best channel quality. We call it best feedback, and this reporting scheme carried out for best four or five among 24 channels has been already adopted in the IEEE 802.16e standard. Considering real feedback conditions, we investigate the performance of a best feedback scheme and derive the optimal number of channels for which information needs to be fed back to keep the throughput gap (compared to a full feedback scheme) within a target margin. From the optimal condition, we propose an adaptive best feedback algorithm, where the number of reported channels is adjusted to adapt to the number of users in the cell. We also propose an adjusted periodic feedback algorithm, where users are divided into groups and scheduling is carried out group by group, so a user can report feedback information when his group is scheduled, thereby reducing the frequency of feedback transmission. To support differentiated performance, we further propose heterogeneous feedback algorithms where users are divided into heterogeneous groups, each with a different group size or a different feedback period. Numerical results validate our analysis and provide meaningful insights into the design of various channel feedback schemes.

Abstract A fully connected one‐hop ad hoc network constitutes a basic unit for managing self‐organizing networks such as IEEE 802.11 and 802.15.3 networks. Since energy efficiency is a critical issue in ad hoc networks, we develop an energy‐saving framework that includes scheduling for node‐to‐node direct communication. The scheduling is performed by a coordinator that is selected by some simple rule. We enhance IEEE 802.11 protocol by using our proposed framework, and analyze its energy efficiency in transmitting and receiving data. Through mathematical analysis, we confirm that our enhanced protocol significantly saves energy compared to the IEEE 802.11 protocol. We also investigate the robustness of our algorithm by covering the cases of uncooperative users, system malfunctioning, and channel errors. The numerical results confirm that our protocol works well under these hostile environments and maintains its advantage over the conventional scheme. Copyright © 2009 John Wiley & Sons, Ltd.

In a cognitive radio (CR) network, cooperative spectrum sensing plays an important role in ensuring the quality of service (QoS) of primary users (PUs) and improving spectral utilization efficiency. Intuitively, the more secondary users (SUs) are involved in sensing, the more sensing accuracy the CR can achieve, whereas the more sensing overhead the SUs consume, the less throughput the CR network can achieve. In this paper, we investigate overhead-throughput tradeoff over fading channels in a cooperative CR network that consists of a number of the SUs employing energy detectors and a single decision fusion center. We propose a design strategy to maximize the throughput of the SU network by choosing appropriate sensing length and the number of the SUs reporting to a decision fusion center. Moreover, we extend our analysis to a two-stage cooperative sensing mechanism where the second-stage fine sensing is triggered whenever any SU reports the presence of a PU after the first-stage detection. Our numerical results show that compared with the single stage sensing, the two-stage sensing scheme achieves higher throughput via the reduction of the probability of false alarm.

The operation of widely-deployed random access to wireless networks is based on limited information on the result of each access attempt. When making a random access attempt, users usually do not know the exact amount of transmit power to make it successful. Also, upon failure of a transmission attempt, a user cannot tell whether the failure was caused by collision with other simultaneously-transmitting users or by his use of insufficient transmit power. To handle lack of information on the cause of failure, we propose an innovative Cause-of-Failure (CoF) resolution which increases the transmit power after a given number of consecutive unsuccessful access attempts when the probability that a given failure is caused by collision becomes sufficiently low. To exploit the thus-achieved transmit power for the next random access attempt, we also determine the Cause-of-Success (CoS) based on the number of consecutive successful attempts, i.e., whether to decrease or maintain the present transmit power probabilistically. This way, users can adjust their transmit power for random access, which we call Auto Power Fallback (APF). We evaluate APF by modeling analysis and numerical computation based on the slotted Aloha, showing that APF makes significant energy-savings for uplink random accesses while achieving good performance.

Emerging systems like OFDMA and MIMO systems require multichannel scheduling over a wireless link. In this paper, we focus on the case that the number of channels to be assigned to a mobile terminal is limited, which we call limited matching. This case often occurs over a MIMO downlink when the number of receive antennas at a mobile user is smaller than that of transmit antennas at the base station. When the system exploits channel-aware opportunistic scheduling based on the channel feedback, finding the sum of channel gains by the limited matching is an NP-complete problem. To solve it easily, we can use the Hungarian algorithm, but its complexity is still too high and not amenable to performance analysis. Hence, we develop a heuristic algorithm, and analyze its cell throughput. Also, we investigate its performance when the channel feedback information is partially available. For analytic simplicity, we consider proportionally fair (PF) scheduling that is widely accepted as an opportunistic scheduler. Numerical results demonstrate that our heuristic limited-matching scheduling algorithm works well with partial channel feedback.

Opportunistic scheduling provides a good chance to improve wireless system performance by exploiting the underlying channel condition. There has been a lot of work on opportunistic scheduling, but the problem of finding the right feedback mechanism to convey channel information has largely been untouched. In emerging multichannel systems, the per- channel feedback induces a substantial amount of feedback overhead and requires high computational complexity. To reduce the feedback overhead, we consider an opportunistic feedback strategy that activates the channel feedback opportunistically according to the channel condition. Then, we combine the opportunistic feedback with the best-n channel feedback scheme where a mobile user chooses the best n channels and transfers this information to the base station. We analyze the throughput and the amount of channel feedback information for proportionally fair opportunistic scheduling under Rayleigh fading i.i.d. channels. The numerical results confirm that our partial feedback schemes achieve a remarkable reduction in the amount of feedback information at the cost of slight throughput degradation, thereby saving the scarce wireless uplink bandwidth and limited battery power.

In this letter, we propose a power-based call admission control (CAC) scheme to accommodate multiclass traffic by directly extending the number-based CAC scheme in multicode CDMA networks, and develop some related mathematical properties. Against the conventional findings, we demonstrate that complete partitioning (CP) of the received signal power at a basestation for each traffic class can be an approach as useful as complete sharing (CS) in accommodating an appropriate number of users. The main advantage of CP scheme over CS scheme is its simplicity in resource management

The OFDMA systems are emerging for future cellular networks. Creating multiple data channels, they can support the flexible frequency reuse factor (FRF). Although FRF 1 is the best choice in terms of cell throughput, it causes intercell interference at the cell boundary, thereby being unable to serve the whole cell area. Therefore it was proposed to use the FRF of greater than 3. In this paper, we develop a flexible FRF design mechanism that provides an intermediate value between 1 and 3 while the conventional schemes are dedicated to use some integer numbers only such as 3, 4, or 7. In our design, if the number of shared channels between any neighboring cells is given, we implement it simply according to a difference set. Simulation results show that a FRF of 7/4 achieves better throughput than FRF 3 and overcomes the intercell interference problem of FRF 1, so it can replace the conventional FRF such as 3. We expect that our new FRFs have the advantage in supporting smooth handoff because there are always some common channels between two neighboring cells.

Opportunistic scheduling can significantly improve wireless network performance by exploiting the feedback information that conveys the underlying channel condition. In emerging multichannel systems, the perchannel feedback induces a substantial amount of feedback overhead and requires high computational complexity. To reduce the feedback overhead, we consider an opportunistic feedback strategy that activates the channel feedback opportunistically according to the channel condition. Then, we combine the opportunistic feedback with the best-n channel feedback scheme where a mobile user chooses the best n channels and transfers this information to the base station. We analyze the throughput and the amount of channel feedback information for proportionally fair opportunistic scheduling under Rayleigh fading i.i.d. channels. The numerical results confirm that our partial feedback schemes achieve a remarkable reduction in the amount of feedback information without a significant throughput degradation, thereby saving the scarce wireless bandwidth and limited battery power.

Orthogonal frequency-division multiple access (OFDMA) systems are considered promising candidates for implementing next-generation wireless communication systems. They provide multiple channels that can be accessed via random access schemes. However, traditional random access schemes could result in an excessive amount of access delay. To address this issue, we develop a fast retrial scheme that is based on slotted Aloha and exploits the structure of OFDMA. A salient feature of this scheme is that when collisions occur instead of retrials occuring randomly in time, they occur randomly in frequency, i.e., the scheme randomly selects the subchannels for retrial. To further achieve fast access, retrials are designed to follow the 1-persistent type, i.e., no exponential backoff. To achieve the maximum throughput, we limit the maximum number of allowed retrials according to the load condition. We also consider the issue of designing for an appropriate reuse factor for random access channels in order to overcome the intercell interference problem in OFDMA multicell environments. Our finding is that full sharing, i.e., a reuse factor of one, performs best for given random access channels. Through analysis and simulation, we confirm that our fast retrial algorithm has the advantage of high throughput and low access delay, and the full sharing policy for random access channels shows high throughput as well as low collision.

As the delay is a critical QoS factor, packet scheduling over a wireless access link that often becomes congested needs to have the objective of meeting each user's delay requirement. To incorporate the delay into the scheduler design, we consider the objective of maximizing the total utility (U_T). However, since a utility-based scheduler that concerns delay requires high complexity, we introduce the concept of marginal utility. Representing the objective as minimizing the total marginal utility (M_T), we develop some related properties for maximizing U_T and minimizing M_T. For the case with fixed service time, we show that the outcome of M_T minimization becomes equivalent to that of U_T maximization. For the more complicated case of varying service time, the M_T minimization sheds light on the design of a simple scheduler. Overall, the marginal utility requires significantly low complexity for packet scheduling compared to the ordinary utility. Through simulations, we confirm that the marginal utility gives a way of flexible scheduling in meeting various delay requirements.

It is well known that opportunistic scheduling by using feedback information significantly improves wireless network performance. Most opportunistic scheduling works have focused on the case where a single channel is shared by multiple users. However, emerging wireless technologies (e.g., MIMO, OFDMA, etc.) are characterized by multiple shared channels, which complicates the problem. Moreover, it is necessary for the network to be able to provide various levels of quality of service (QoS). To address these issues, we develop a QoS-aware selective feedback model and a method to do optimal resource allocation. In our feedback model, each user chooses those channel sets that meet its QoS requirements by exploiting user diversity, thus resulting in a significant reduction in the amount of feedback information. Given the feedback channel sets for each user, the base station then distributes channels to each user with the objective of maximizing the number of accommodated users or the sum of users' utility values. We use a graph theoretic approach to solve these maximization problems by mapping them to clique searching problems. We develop some interesting theoretical results and properties but show that the complexity of this problem can be exponential in the number of channels. Thus, we also develop two suboptimal algorithms to handle the case when the number of shared channels is large. Finally, we demonstrate the efficacy of our results through an extensive numerical study

In this paper, we suggest a framework for power saving (PS) MAC protocol in a one hop mobile ad hoc network with a coordinator, and propose affinity based scheduling algorithm. Affinity is a novel concept that we introduce first. We define two types of affinities, i.e., node affinity and task affinity. Node affinity is defined as the weighted sum of incoming and outgoing task flows at each node, and task affinity as the sum of node affinities of source and destination nodes of each task. Our proposed scheduling algorithm has the advantage of putting more nodes into doze state for energy saving after completion of each task. To maximize the PS effect, we combine shortest job first scheduling with affinity based scheduling. We also devise an enhanced 802.11 PS MAC protocol by applying our framework and scheduling algorithm to the 802.11 MAC protocol which prevails in real world. Simulation results demonstrate that our scheduling algorithm and PS MAC protocol improve power efficiency and throughput significantly over other competitive schemes.

IEEE 802.11 wireless LANs are evolving into a high speed system by adopting MIMO technologies. Currently the standard deals with transmission for one user over a link. In this paper, we consider the MIMO technique of spatial multiplexing that enables multiple users to receive packets over the downlink simultaneously. It takes advantage of multiuser diversity in the space and time domains, supposing that each antenna independently performs link adaptation for each subchannel. Through analysis, we show that multiuser transmission has performance improvement over the single-user case. However, by using adaptive data splitting over spatial multiplexing, the performance of the single-user transmission is enhanced, and even better than that of multiuser case for small multiuser diversity. To exploit the multiuser diversity extensively, we apply a scheduling algorithm that considers channel condition of each subchannel. Simulation results show the multiuser transmission has higher link utilization than the single-user case for large multiuser diversity. To implement this, the system requires further considerations to modify ACK policy, and to mix conventional and MIMO-capable stations.

A bandwidth-limited wireless channel can considerably improve its performance by exploiting multiple-input-multiple-output (MIMO) antennas. Combining spatial multiplexing with multiuser diversity, we develop an optimal cross-layer scheduling mechanism that executes fair scheduling at the upper layer and optimal antenna assignment at the physical layer. For fair scheduling, we propose a framework that achieves the objective of maximum capacity and proportional fairness. For optimal antenna assignment, we consider the Hungarian algorithm that maximally utilizes the characteristics of MIMO systems by adopting the graph theoretical approach. Through simulations, we demonstrate the performance of the optimal scheduling.

Recently cdma2000 1x EV-DO (HDR) system has begun to be deployed in some countries to support high data rate services in cellular networks. The system is originally designed to support data services, but now is expected to serve some real-time traffic including VoIP. For VoIP service with delay hound and low loss requirements, we propose a frame structure considering delay bound and a scheduling algorithm reflecting channel conditions. To schedule VoIP, we adopt the maximal rate algorithm and the proportionally fair algorithm. The proportionally fair algorithm (PF) was known to be appropriate for elastic-traffic, however, from simulation results, we conclude that the PF algorithm with the channel test is an appropriate scheduling scheme to provide QoS of VoIP. When the required slot portion of VoIP is 75%, the loss rate is about 1% on the average and 3% in the worse case. On the other hand, the maximal rate algorithm shows twice of the loss rate for the same delay bound and load. Additionally we propose a simple admission control scheme for VoIP service that controls the average portion of slots occupied by VoIP packets.

Exploiting multiple-input-multiple-output (MIMO) diversity, systems with error-prone and bandwidth-limited wireless channels can easily support reliable transmission. We develop optimal cross-layer scheduling that consists of QoS scheduling at the upper layer and optimal antenna selection at the physical layer. For QoS scheduling, we design a framework for optimal scheduling to meet users' QoS requirements. To solve this optimization problem, we consider a clique-searching algorithm for antenna selection that maximally utilizes the characteristics of MIMO systems by adopting the graph theoretical approach. As the clique searching problem becomes NP complete with the increase of transmit antennas in number, we propose a suboptimal antenna selection algorithm to deal with a large number of transmit antennas. We derive some theorems and properties for our approach and, through simulations, we demonstrate the performance of QoS scheduling which is effective to handle real-time traffic.

CDMA data networks such as cdma2000 1x EVDO are proposed in the midst of evolving to the 3rd generation wireless networks. Basically they use time division multiplexing and rate control that need a downlink scheduling to increase the system capacity, thereby being able to support high speed data rates. As the systems will eventually support multimedia and data traffic together, we need to have a propose criterion for scheduling that can count various service requirements such as delay and loss. Therefore, we visit the concept of utility and opportunity cost considering these together. The opportunity cost is defined as the maximum utility lost among the other users by giving the current turn to a particular user. We design an algorithm to select a job for transmission with the maximum profit that is obtained by subtracting the opportunity cost form its expected utility. The simulation results show that it can support various QoS levels in terms of delay and loss for various traffic scenarios.

Cellular systems are designed to support voice service over the wireless channel but nowadays data-only networks such as cdma2000 1x EV-DO and high speed downlink packet access (HSDPA) are being deployed in the midst of evolution to the next generation wireless networks. Such systems become important because data services are essential in future wireless networks that are supposed to adopt the concept of all-IP to support integrated services. To increase the data transmission rate at downlink in cdma2000 1x EV-DO and HSDPA, the base station (BS) uses only a channel multiplexed by time division, not by code division. These systems use proportional fair scheduling to maximize throughput. While it is useful for non real time traffic, it does not provide appropriate QoS for real time services such as voice and video. Therefore we need to incorporate the multimedia QoS requirements into the design of a new scheduling algorithm. Our proposed wireless-adaptive fair scheduling (WAF) tries to allocate time slots with fair share according to the stream requirements considering the varying channel conditions to achieve good throughput. The simulation results show that our scheme can guarantee fair service and increase throughput by up to 41% compared to the purely fair (PF) scheduling scheme.

In multiple cellular CDMA systems, the channel capacity heavily depends upon the interference caused by the signal power of the other users. In this paper, we exactly quantify the cell capacity by focusing on not only the number of users but also the location of active users. To inspect the effect of multicell CDMA environments, we model three 1/6 cells divided by 60/spl deg/ sector antennas and analyzed the increased interference when a user is added to the unit of cell. Based on the capacity analysis, we propose a new call admission control scheme that uses the location information of users. Whenever a new user arrives, our scheme checks whether the admittance of that user can occur within the cell capacity. Our scheme allows more users to enter the network if they generate less interference. Therefore it has the effect of increased system capacity without sacrificing the QoS of all the other active users. Through simulations we show that the performance is improved by 10 to 20% in terms of the number of active users that can be accommodated.