The evolution of 5G technology towards 5G-Advanced has introduced advanced vehicular applications with stringent Quality-of-Service (QoS) requirements. Addressing these demands necessitates intelligent resource management within the standard 3GPP network slicing framework. This paper proposes a novel resource management scheme leveraging a Deep Deterministic Policy Gradient (DDPG) algorithm implemented in the Network Slice Subnet Management Function (NSSMF). The scheme dynamically allocates resources to network slices based on real-time traffic demands while maintaining compatibility with existing infrastructure, ensuring cost-effectiveness. The proposed framework features a two-level architecture: the gNodeB optimizes slice-level resource allocation at the upper level, and vehicles reserve resources dynamically at the lower level using the 3GPP Semi-Persistent Scheduling (SPS) mechanism. Evaluation in a realistic, trace-based vehicular environment demonstrates the scheme’s superiority over traditional approaches, achieving higher Packet Delivery Ratio (PDR), improved Spectral Efficiency (SE), and adaptability under varying vehicular densities. These results underscore the potential of the proposed solution in meeting the QoS demands of critical 5G-Advanced vehicular applications.

Large language model (LLM) inference poses dual challenges, demanding substantial memory bandwidth and computing resources. Recent advancements in near-memory accelerators leveraging 3D DRAM-to-logic hybrid-bonding (HB) interconnects have gained attention due to their highly parallel data transfer capabilities. We address limitations in previous HB-DRAM accelerators, such as those stemming from distributed controller designs, by introducing an architecture with a centralized controller and dual-IO scheme. This approach not only reduces the chip area overhead but also enables reconfigurable GEMV/GEMM operations, boosting the performance. Simulations for the OPT 66B model show that our proposed accelerator achieves 2.9X, 3.5X, and 2.5X higher performance compared to NPU, DRAM-PIM, and heterogeneous designs (DRAM-PIM + NPU), respectively.

NR-V2X Mode 2 is introduced by the third generation partnership project (3GPP) to support vehicle-to-everything (V2X) communication. In NR-V2X Mode 2, vehicles select resources for the exchange of cooperative awareness messages (CAM) in a decentralized manner based on their local observation using semi-persistent scheduling. Resources are distributed over the 2-D frequency and time domain, following the long-term evolution frame structure. Since vehicles select resources based on their local observations and due to spectrum scarcity, this may lead to contention. Hence, selecting a resource is challenging, and as each vehicle strives to select a resource, it becomes a consensus problem. To resolve resource contention, in this article, we propose a knowledge-empowered distributed multiagent deep reinforcement learning (K-MADRL) approach. Based on traffic flow information, long short-term memory (LSTM) is employed to deploy Unmanned Internet of Aerial Agents (UIAAs) to collect vehicle state information. UIAAs gather vehicle state knowledge and train the local deep reinforcement learning (DRL) model. The locally trained model at the UIAA is shared and aggregated at the gNB for the global model update. The trained policy is then sent to the vehicles over system synchronization blocks for distributed execution. Moreover, the vehicles select the resource based on the joint action, i.e., by anticipating the actions of the neighboring vehicles. Our scheme is compared with other methods, such as DRL, optimization techniques, the SPS method, and random allocation methods, used in the NR-V2X environment. The results of the simulations show that our scheme outperforms the other methods.

The Third Generation Partnership Project (3GPP) introduced cellular vehicle-to-everything (C-V2X) for vehicular communications. In the standard, C-V2X Mode 4 is defined for the distributed resource selection. Subsequently, in 3GPP Release 16, NR-V2X is introduced with Mode 1 and Mode 2 for vehicular communications. Likewise C-V2X Mode 4, NR-V2X Mode 2 is used for decentralized resource scheduling. The vehicles select the resources based on their local observations by utilizing the semi-persistent scheduling (SPS). Since, the vehicles select the resources based on the local observation, sensing nature of SPS is challenged by the hidden node problem that lead to resource conflict. To resolve the contention, 3GPP also introduced the physical sidelink feedback channel (PSFCH) to assist the distributive resource scheduling based on the receiver feedback. However, this incurred a signaling overhead. In this work, federated learning is exploited for distributive training via offline method and distributive multiagent-based resource scheduling is performed following the principles of NR-V2X Mode 2. Distributed training favors the model accuracy by accommodating the varying affect of the environment due to the high mobile dynamics. Simulation is conducted by integrating SUMO in conjunction with 3GPP NR-V2X standard. Performance results demonstrate a substantial improvement compared to other deep learning methods, where centralized training and random resource selection procedures are employed. This research marks a significant stride toward efficient and conflict-resilient resource allocation in vehicular communications.

Nowcasting plays a crucial role in responding to disasters such as flash floods by predicting rainfall in real time. However, existing nowcasting models struggle to accurately predict extreme rainfall events, which, although rare, can have devastating impacts. This challenge primarily arises because typical loss functions focus on minimizing average prediction errors rather than emphasizing the importance of extreme events, leading to their underestimation. To address this issue, this study introduces various weighted loss functions that impose greater penalties on prediction errors as rainfall intensity increases. These weighted loss functions were applied to a convolutional long short-term memory (ConvLSTM)-based nowcasting model to assess their impact on model performance. Recognizing that weighted loss functions may influence the learning of spatial patterns, and this study categorized extreme rainfall events based on their spatial characteristics and conducted a detailed performance evaluation for each type. The results demonstrated that models using weighted loss functions significantly improved the accuracy of extreme rainfall predictions compared to unweighted (UW) models. Notably, depending on the applied weighted loss functions, each model clearly exhibited its strengths and weaknesses across various extreme rainfall types. This finding suggests that selecting the best-performing weighted model based on prediction goals can lead to optimal results. Furthermore, this study revealed that the effectiveness of prediction methods varies significantly depending on the type of extreme rainfall event, indicating the need for dynamic selection of prediction methods tailored to specific condition. This article provides valuable insights into improving extreme rainfall nowcasting and is expected to contribute to enhancing disaster response systems in the future.

When the training dataset comprises a 1:1 proportion of dogs to cats, a generative model that produces 1:1 dogs and cats better resembles the training species distribution than another model with 3:1 dogs and cats. Can we capture this phenomenon using existing metrics? Unfortunately, we cannot, because these metrics do not provide any interpretability beyond “diversity". In this context, we propose a new evaluation protocol that measures the divergence of a set of generated images from the training set regarding the distribution of attribute strengths as follows. Singleattribute Divergence (SaD) reveals the attributes that are generated excessively or insufficiently by measuring the divergence of PDFs of individual attributes. Paired-attribute Divergence (PaD) reveals such pairs of attributes by measuring the divergence of joint PDFs of pairs of attributes. For measuring the attribute strengths of an image, we propose Heterogeneous CLIPScore (HCS) which measures the cosine similarity between image and text vectors with heterogeneous initial points. With SaD and PaD, we reveal the following about existing generative models. ProjectedGAN generates implausible attribute relationships such as baby with beard even though it has competitive scores of existing metrics. Diffusion models struggle to capture diverse colors in the datasets. The larger sampling timesteps of the latent diffusion model generate the more minor objects including earrings and necklace. Stable Diffusion v1.5 better captures the attributes than v2.1. Our metrics lay a foundation for explainable evaluations of generative models.

Network slicing, a key technology of next-generation wireless networks, has undergone significant evolution from its inception as Dedicated Core Network (DCN) in 4G-LTE to its current state in 5G-Advanced. This paper provides a comprehensive analysis of network slicing enhancements across 3GPP releases 13 to 17, categorized into three phases: 5G-Basic (Release 15), early 5G-Evolution (Release 16), and advanced 5G-Evolution (Release 17). Furthermore, our study identifies persistent challenges in network slicing implementation and proposes innovative enhancements for 5G-Advanced (Release 18), including a novel machine learning-based approach to minimize service interruptions within a Registration Area (RA). This approach combines predictive insights from a Long Short-Term Memory (LSTM) model with a Dynamic Proportional Resource Allocation (DPRA) method for resource reconfiguration. Evaluation of the LSTM-DPRA scheme demonstrates significant performance improvements and reduced service interruptions compared to benchmark schemes, contributing to the development of more efficient and reliable network slicing.

The precise correction of outliers within radar rainfall data is crucial for a wide range of applications, including the analysis of extreme rainfall events, hydrological modeling, and the forecasting and warning of flash floods. Despite its significance, the challenge of correcting these outliers has not yet been fully explored, mainly due to the high dimensionality and spatiotemporal intricacies of radar rainfall data. Furthermore, most existing techniques for outlier correction are overly simplistic, revealing limitations when it comes to effectively correcting sporadic outliers. In response, this study has developed a novel approach of detecting and correcting outliers based on radar rainfall statistics at local spatiotemporal scale. In this approach, an algorithm of detecting outliers based on the simple 3-sigma rule in spatiotemporal context and an algorithm of detecting abrupt change between adjacent radar cells in spatial context, all in local scale, are iterated to enhance the quality of radar rainfall data progressively and effectively. This correction method resulted in a radar rainfall data with the grid cell value closely resembling that of the ground gauge data as well as the probability distribution. In addition, when compared to the existing methods, it demonstrated its ability to selectively remove only the outliers while preserving the integrity of the normal data. What sets this proposed method apart is not only its practicality, as it relies solely on 2D radar reflectivity data and can be easily implemented, but also its contribution to improving the analysis accuracy across various domains reliant on radar rainfall data.

Content retrieval in content-centric vehicular networks faces challenges that include high latency, especially when content is stored far from the requesting vehicle. On-path caching feature in the conventional vehicular named data Networks (VNDN) enables content storage that can reduce latency. However, due to the constantly changing dynamic ad hoc nature of the vehicular network, the availability of stored content for the requester vehicle cannot be guaranteed. In addition, without knowing which content will be requested, where it will be requested and when it will be requested, the content caching functionality of VNDN is underutilized. To address this issue, this manuscript proposes a content prefetching scheme for the Content-centric Internet of Vehicles (CIoV) by introducing the content Value of Popularity ($VoP$) matrix. Considering vehicles requesting content of similar interests, we evaluate$VoP$through three value update functions that follow the power law of the time elapsed since the last content requested. By multiple parameters of consumer vehicle similarity, an on-road proactive content retriever vehicle is selected. The simulation results showed that the proposed proactive on-path content prefetching mechanism significantly reduces the content delivery delay while increasing the success delivery ratio by 48% and extends the spread of content within the network by 53%.

Data provenance, or data lineage, describes the life cycle of data. In scientific workflows on HPC systems, scientists often seek diverse provenance (e.g., origins of data products, usage patterns of datasets). Unfortunately, existing provenance solutions cannot address the challenges due to their incompatible provenance models and/or system implementations. In this paper, we analyze four representative scientific workflows in collaboration with the domain scientists to identify concrete provenance needs. Based on the first-hand analysis, we propose a provenance framework called PROV-IO$^+$, which includes an I/O-centric provenance model for describing scientific data and the associated I/O operations and environments precisely. Moreover, we build a prototype of PROV-IO$^+$to enable end-to-end provenance support on real HPC systems with little manual effort. The PROV-IO$^+$framework can support both containerized and non-containerized workflows on different HPC platforms with flexibility in selecting various classes of provenance. Our experiments with realistic workflows show that PROV-IO$^+$can address the provenance needs of the domain scientists effectively with reasonable performance (e.g., less than 3.5% tracking overhead for most experiments). Moreover, PROV-IO$^+$outperforms a state-of-the-art system (i.e., ProvLake) in our experiments.

Wind power generation plays an increasingly significant role in the global shift towards renewable energy sources for climate mitigation. However, its susceptibility to climate variability underscores the critical importance of accurate energy generation prediction for ensuring a stable energy supply. In this paper, we analyze the correlation between climate data and energy generation data, extracting essential factors based on this analysis. We propose a Convolutional Neural Network-based model capable of four-hour short-term forecasting by representing these factors as embedding matrices. Furthermore, we combine this model with a Long Short Term Memory model to extend the forecasting period to 24 hours, validating its performance in the day-ahead market bidding context. Using empirical data from South Korea, our short-term forecasting model achieved an accuracy of 76%, while the long-term model demonstrated 85% accuracy, highlighting its potential for practical applications in wind energy generation and market operations.

This study developed a Deep Neural Network (DNN) based distributed hydrologic model for an urban watershed in Republic of Korea. The developed model is composed of multiple Long Short-Term Memory (LSTM) hidden units connected by a fully connected layer. To examine the study area using the developed model, time series of 10-minute radar-gauge composite precipitation data and 10-minute temperature data at 239 model grid cells with 1km resolution were used as inputs to simulate 10-minute watershed flow discharge as an output. The model performed well for the calibration period (2013-2016) and validation period (2017-2019), with Nash-Sutcliffe Efficiency coefficient values being 0.99 and 0.67, respectively. Further in-depth analyses were performed to derive the following conclusions: (1) the map of runoff-precipitation ratios produced using the developed DNN model resembled imperviousness ratio map of the study area from the land cover data, revealing that the DNN successfully deep-learned the precipitation partitioning processes only with the input and output data without depending on any priori information about hydrology; (2) the model successfully reproduced the soil moisture dependent runoff process, an essential prerequisite of continuous hydrologic models; (3) each LSTM unit has different temporal sensitivity to the precipitation stimulus, with fast-response LSTM units having greater output weight factors near the watershed outlet, which implies that the developed model has a mechanism to separately consider the hydrological components with distinct response time such as direct runoff and the groundwater-driven baseflow.

The advancement of wireless sensor networks (WSNs) improves various smart home automation services and home users’ living standards. However, efficiently collecting data and automating smart home services require the extensive deployment of the sensors. Thus, one of the crucial and challenging tasks is to minimize the sensors’ energy consumption for monitoring and automating various activities in a smart home. In this article, we present a solution to control the excessive energy consumption of sensors used to detect various activities of daily living (ADL) of a smart home resident. The sensors within a smart home network are divided into various groups employing the recurrent neural network (RNN) and dynamic time warping (DTW) techniques to predict the activities with high accuracy and less energy consumption. The smart home users’ future activities are forecast with bidirectional long short-term memory (BLSTM) RNN model to select those sensors that are likely to predict the upcoming activities. Similarly, to predict the home users’ unusual activities, a guard sensor is elected among sensors with high similarities with each other using DTW. The sensor’s role is evenly switched between different modes to maintain a fair tradeoff between energy and accuracy. An extensive set of simulations is performed to validate the proposed scheme’s work integrating datasets from authentic sources. Finally, the proposed system significantly reduces the sensors’ energy consumption and prolongs the battery lifetime to approximately 137 days.Note to Practitioners—This article presented an energy-efficient duty-cycling scheme for automating sensors’ operations in a smart home scenario. The traditional duty-cycling schemes mainly provide solutions based on appointing sentries or predicting smart home users’ next activity using models, such as Bayesian networks. We design a system that integrates the advantages of both sentry and prediction-based schemes to reduce the amount of energy required by sensors to detect and automate smart home users’ activities with high accuracy and precision. The active sensors are appointed using a bidirectional long short-term memory recurrent neural network. Similarly, the guard sensors are assigned to detect unusual activities using the similarities among idle sensors. This study could be used to automate the smart home sensors for detecting home user’s activities with less energy, which ultimately prolongs the battery lifetime of the sensors.

Day by day innovation in wireless communications and micro-technology has evolved in the development of wireless sensor networks. This technology has applications such as healthcare supervision, home security, battlefield surveillance and many more. However, due to the use of small batteries with low power this technology faces the issue of power and target monitoring. There is much research done to overcome these issues with the development of different architecture and algorithms. In this paper, a scheduling machine learning algorithm called adaptive learning automata algorithm(ALAA) is used. It provides an efficient scheduling technique. Such that each sensor node in the network has been equipped with learning automata, and with this, they can select their proper state at any given time. The state of the sensor is either active or sleep. For the experiment, different parameters are used to check the consistency of the algorithm to schedule the sensor node such that it can cover all the targets with the use of less power. The results obtained from the experiments show that the proposed algorithm is an efficient way to schedule the sensor nodes to monitor all the targets with use of less power. On the whole, this paper manages to achieve its goal by contributing to the related research on wireless sensor networks with a new design of a learning automata scheduling algorithm. The ability of this proposed algorithm to use the minimum number of sensors to be in active state verified to reduce the use of power in the network. Thus, achieving the goal by enhancing the lifetime of wireless sensor networks.

In underwater wireless sensor networks (UWSN), acoustic communication naturally introduces challenges such as long propagation delay and high packet loss. The flooding-based routing protocol can address these challenges with its multipath characteristics. As in flooding-based routing, due to multipath propagation mechanism, not only DATA but also ACK messages are transmitted through multiple routes however still some packet loss will degrade the performance. So, to provide high reliability of message delivery, an efficient retransmission mechanism is inevitable. Though, if the network uses conventional transport layer protocol such as TCP, it will suffer a spurious retransmission problem as TCP was originally not designed for the multipath environment. In this paper, we propose route discrimination for flooding-based routing to reduce spurious retransmission in UWSN to solve the limitation. The notion of ACK copies waiting time (ACWT) is utilized which is selectively updated based on the similarity of paths of transmission of ACK message copies. We also improved our previous solution that lacks flexibility to cope with dynamic link error characteristics. Through evaluation, we verified that our new scheme achieves the performance improvements of 14%~84% in terms of retransmission ratio compared to the previous research.

In Underwater Wireless Sensor Networks (UWSN), Autonomous Underwater Vehicles (AUVs), which are responsible to collect(deliver) data from(to) sensors, bridge the network to the rest of the world-wide network to form the Internet of Underwater Things (IoUT). In IoUT acoustic sensors that cause long propagation delay together with slow AUV speed could render the collected data useless if not delivered to the sink in timely manner. In addition, in security, environment monitoring, and emergency applications the age of the data cached by the AUV is very crucial. In this paper, we harnessed the end-to-end data freshness constraint to design AUV path of traversal. The resultant dynamic path of traversal improves the overall data freshness of the sectors of UWSN. In the evaluation, compared to the conventional lawnmower and shortest path traversal algorithms, the proposed scheme improved the overall data freshness at the cost of data collection delay.

The unique characteristics of underwater wireless sensor networks (UWSN) pose many constraints in transmission of data packets and energy efficiency is one of them. The battery powered underwater acoustic nodes require sophisticated protocols to control the nodes' active and sleep periods in order to increase the nodes lifetime. In the terrestrial networks a number of variants of Receiver Initiated MAC (RI-MAC) protocols utilize the nodes active and sleep cycles with the aim to improve the energy efficiency of the overall network. Motivated by the significance of less costly and energy efficient RI-MAC protocol, we present its two variants for the UWSN. In the first protocol, Duty Cycle Scheduling based on Residual Energy (RidE), each sender node to adjust its duty-cycle based on its residual energy. In the second protocol, Duty Cycle scheduling based on Next Wake-up Time (NeWT), each node calculates its duty cycle according to the duty cycle of last awake node, this allows the nodes to avoid any data collision and utilize their sleep modes efficiently to conserve their energy. The simulation results show that, RidE alleviates the need of additional re-transmissions as all the sender nodes overhear the communication with the receiver and plan their communication accordingly at the cost of additional energy consumption. Whereas, in NeWT energy consumption is minimized as the nodes stay awake for shorter period.

Recently, Underwater Wireless Sensor Networks (UWSN) have been proven to provide numerous application including military, environmental, and pollution surveillance, etc. However, once sensors are deployed in the deep sea, it is merely possible to recharge or replace their batteries. Hence, an efficient management of the available resources can extend the network lifetime. For example, clustering the sensor nodes is a potential solution, yet, exchange of multiple packets for cluster head selection, packet collision while sending data to cluster head and the continuous awakening of nodes waste a considerable amount of energy. Utilizing the advanced technology of the Autonomous Unmanned Vehicle(AUV) in UWSN, we present AUV assisted Energy-efficient Clustering(AEC) mechanism that introduces wake-up sleep cycle for the underwater sensor nodes. In the designed scheme, AUV not only collects data but also operates as a central regulator. AUV onus includes cluster creation, the cluster head nomination and creation of a wakeup- sleep schedule for the UWSN that relieves the additional burden from energy limited underwater sensor nodes. Unlike traditional clustering mechanisms, no additional packets exchanged in energy-efficient dynamic cluster head selection. The proposed scheme is evaluated and compared with Clustering with Fixed Cluster head(CFC) scheme. Simulation results show that by applying AEC, the network can remain stable for a long time that allows UWSN to deliver data reliably for the extended time span.

In this paper, we propose a distributed interest forwarder selection (DIFS) scheme that mitigates the interest broadcast storm in vehicular named data networks. In DIFS, a vehicle sends an interest packet piggybacking its location, distance to the neighbors, and speed. In this case, the immediate neighbors do not have the requested content and rank themselves to be an eligible interest forwarder by using multiple attributes. Additionally, every intermediate vehicle uses a digital map to be selected as forwarders in both (forward and backward) directions of the consumer. Simulations show that DIFS satisfies more interest packets with less delay as compared with the recent forwarding solutions.

In Vehicular Delay Tolerant Networks, a number of Roadside Units (RSUs) are deployed along the road and connected to the infrastructure network to provide various services to the vehicles on the road. However, it is hard to cover the long highways completely, due to the deployment cost. In such uncovered areas between two neighboring RSUs, a connection between a vehicle and an RSU cannot be established. To cope with this, few schemes have been proposed recently, enabling one RSU to select one relay vehicle to provide continuous communications for the vehicle moving in the uncovered area. However, the selection of an appropriate relay vehicle for pre-storing maximum data is an open issue. In this paper, we, therefore, propose an adaptive multiple-relay selection scheme that allows RSU to select relay vehicles while taking most relevant multiple criteria into the account. The relay selection triggers when a vehicle is unable to receive all the requested data from the corresponding RSU. The simulation results show that our scheme enables vehicles to retrieve maximum amount of the requested data in uncovered areas.

During the past decades, we have witnessed significant research and development efforts related to Sensor Networks (SN) with both wired and wireless technologies. Those efforts resulted in the development of benchmark application paradigms for SNs with applications in both safety-critical and non-safety-critical fields, including environmental monitoring, military apps, tactics, smart sensing, underwater sensing, pollution sensing, fire alarms, and other hundreds of applications to improve our lifestyles. Similarly, the Internet of Things (IoT) has been promoting the emergence of various networking technologies with standardization efforts being carried out by the IEEE community. SNs have been proven to be the most investigated topic for enabling IoT applications. For example, sensors installed in a house may update the owner on his/her cell phone about any medical or security threat. However, to support such applications, SNs and IoT come up against several constraints, like low capacity processing capabilities, battery operated devices, limited transmission ranges, and limited data transmission capacity. One of the key issues in IoT and SNs is the efficient data forwarding in terms of energy consumption, multi-hop data retrieval, safety/emergency message dissemination, etc. The goal of this special issue is to present and highlight the novel advances, emerging technologies, and applications targeting the efficient data collection is the internet of things and sensor networks. This special issue contains thirteen regular papers which can be organized into three groups: data forwarding in SNs, energy efficiency in wireless body area networks, secure data forwarding in SNs.

In deep waters, both the natural acoustic systems (such as marine mammals) and artificial acoustic systems (like underwater sensor networks (UWSNs) and sonar users) use acoustic signals for communication, echolocation, sensing, and detection. This makes the channel spectrum, heavily shared by UWSNs posing several salient features such as narrow bandwidth, long propagation delay, and high packet loss caused by acoustic channel. Flooding of data packets in such environment, therefore, is known to be a more appropriate mechanism. Hence, many researchers proposed flooding-based routing protocols for UWSNs such as VBF and HH-VBF. Nevertheless, these known protocols maximize overhead within network due to their dependency on additional parameters such as routing vector. To deviate this overhead and control the flooding, we previously proposed a Directional Flooding Routing (DFR) protocol, which controls the flooding area based on link quality. However, DFR lacks to deal with dynamic changes under the shores due to fixed system parameters throughout the network communications. In this paper, we therefore propose two new DFR protocols, called QA_DFR_AA (QoS-Aware DFR with Angle Adaption) and QA_DFR_TA (QoS-Aware DFR with Threshold Adaption) to reflect QoS dynamically. In addition, we also aid our current DFR with holding time technique to avoid packet collision and redundant packet transmission. Through NS-2 simulations, we found that our new intelligent DFR (iDFR) and two new versions of DFR outperforms the current version of the DFR.

In Vehicular Delay Tolerant Networks (VDTNs), a number of Roadside Units (RSUs) are deployed along the road and connected to the infrastructure network to provide various services to the vehicles on the road. However, it is hard to cover the long highways completely, due to the deployment cost. In such uncovered areas between two neighboring RSUs, a connection between a vehicle and an RSU cannot be established. To cope with this, few schemes have been proposed recently, enabling one RSU to select one relay vehicle to provide continuous communications for the vehicle moving in the uncovered area. However, the selection of an appropriate relay vehicle for pre-storing maximum data is an open issue. In this paper, we, therefore, propose an adaptive multiple-relay selection scheme that allows RSU to select relay vehicles while taking most relevant multiple criteria into the account. The relay selection is initiated when a vehicle is unable to receive all the data that is requested in the coverage of the RSU. The simulation results show that our scheme enables vehicles to retrieve maximum amount of the requested data in uncovered areas.

The IEEE 1609.4 standard allows a single radio device to utilize multiple channels by alternating control channel (CCH) and service channel (SCH). During the SCH interval, the RTS/CTS/data/ACK handshake can be triggered to transmit large size of data without the hidden node problem. However, it can cause the exposed node problem that hinders concurrent transmissions, which is fatal in highly dynamic VANETs. Even though judicious SCH selection in a multi-channel environment can mitigate the exposed node problem, IEEE 1609.4 does not specify how to select a SCH, which can cause the randomly selected SCHs to be biased. Conforming to the current standards, we therefore propose a novel scheme that enables the exposed vehicles to avoid selecting the same SCH by piggybacking a candidate SCH selection within the optional field of the basic safety message. Through extensive simulations, it is verified that the average throughput can be improved by up to 26%.

Named Data Networking (NDN) has been recently added to the future Internet family. NDN is basically an extension to the Content Centric Network (CCN) and is expected to support various applications. Those applications are to be supported by the future internet architectures. NDN believes in naming the content rather than using end-to-end device names. Recently, NDN has been adapted into Vehicular Ad hoc Networks (VANETs) and hence, we name it Vehicular NDN (VNDN). At it's early stage, VNDN faces several challenges such as consumer/provider mobility, Interest/Data forwarding, content caching and so on. Mostly, VNDN relies on the fact that Data is sent back to the consumer via same path the Interest packet was received from. However, we analyzed that it's not true in a VANET and there is lack of discussion about managing the Data flow back to the consumers in the current literature of VNDN. In this paper, we therefore, pursue to control the data flooding/broadcast storm of the conventional VNDN by proposing our scheme “CONET”. The main idea of CONET is to allow the consumer vehicle to start hop counter in the Interest message and upon receiving that interest by any potential provider, to include Time To Live (TTL) value with data messages. The TTL value includes the number of hops, Data packets should travel on its way back to the consumer. Simulation results show that CONET forwards less Copies of Data Messages Processed (CDMP) while achieving similar Interest Satisfaction Rate (ISR) as the basic VNDN. In addition, CONET also minimizes the overall Interest Satisfaction Delay (ISD), respectively.

The current IEEE 1609.4 standard defines multi-channel operations to alternate control and service channel intervals during a period of 100ms. However, there is no mention of service channel selection for a service provider, which allows hidden service providers to select the same service channel. This limitation can cause the hidden terminal problem during the service channel intervals, leading to significant performance deterioration. Without modifying the existing standards, our proposed scheme enables hidden service providers to avoid selecting the same service channel by delivering their candidate service channel number in the optional field of the basic safety message (BSM). Through extensive simulations, it is verified that the packet reception ratio can be improved by up to 23% in typical broadcast scenarios.

Recently, Vehicular Ad Hoc Networks (VANETs) have been providing a number of services for on-road users, including video content retrieval using 3G and 4G networks. However, owing to the highly dynamic network topology of VANETs, such services is highly susceptible to poor performance. In such a real-time scenario, a requesting vehicle may not be able to guarantee the video quality using its own wireless interface independently. Thus, motivated to provide a quality video stream, we propose a Fleet Based video Retrieval (FBR) scheme that allows the requesting vehicle to download the H.264/SVC encoded video stream in collaboration with its 1-hop neighbors. The collaborators are selected considering the multiple characteristics of each neighboring vehicle, i.e., 1-hop distance, link duration, velocity, and the available cellular bandwidth. The selected vehicles download the video data using their wireless link and then forward it to the requested vehicle through a Dedicated Short-Range Communication (DSRC) protocol. For comparison, we first evaluate our FBR with 3G equipped vehicles and secondly, we designed FBR to work in 4G/LTE environment. Through simulations, we found that FBR has outperformed the recently proposed scheme in terms of receiving video quality and video-flow handling.

Recently, the Delay Tolerant Networks (DTN) have been utilized in various operational communication paradigms. This includes the communication scenarios that are subject to disruption and disconnection as well as the scenarios with high delay and frequent partitioning, i.e., Vehicular Ad hoc Networks (VANETs). Due to several characteristics match, a new research paradigm named as Vehicular Delay Tolerant Network (VDTN) is introduced. Through relays and store-carry-forward mechanisms, messages in VDTNs can be delivered to the destination without an end-to-end connection for delay-tolerant applications. However, the choice of routing algorithms in VDTNs is still under study. Numerous routing protocols have been proposed to meet requirements of many applications. In this paper, we therefore provide some detailed study of recently proposed routing schemes for VDTNs. We also perform comparative analysis on the basis of unique criterion such as forwarding metrics with their implementations. In addition, open challenges and future directions are provided to make room of interest for the research community.

Many safety applications rely on multi-hop broadcasting to disseminate safety messages. In most existing multi-hop broadcasting protocols, one next forwarder is selected through contention among forwarder candidates based on their different waiting times. In this paper, we first analyze the latency and collision of the existing protocols, and point out two problems: 1) unnecessary delay occurs in the contention process due to the lack of considering the distribution of vehicles and 2) the short difference between waiting times of forwarder candidates may allow redundant broadcasts to collide with each other. Secondly, we propose a new multi-hop broadcast protocol called RObust and Fast Forwarding (ROFF) to mitigate both problems. ROFF solves the first problem of unnecessary delay by allowing a forwarder candidate to use the waiting time which is inversely proportional to its forwarding priority. A forwarder candidate acquires its forwarding priority using the novel concept of ESD bitmap, which describes the distribution of empty spaces between vehicles. In addition, ROFF prevents the waiting time difference from being shorter than the predefined lower bound in order to avoid collisions, thus solving the second problem. Our extensive simulations reveal that ROFF achieves faster and more reliable broadcasting as compared to the other protocols.

In duty-cycled wireless sensor networks, senders suffer from long delays while waiting in awaken state for receivers to wakeup for packet transmissions. It results in unnecessary energy usages and lower network lifetime. Opportunistic packet forwarding schemes have been proposed to reduce such delays where the selection of next hop forwarder is performed opportunistically without waiting for a particular neighbor to wakeup. Even under the opportunistic nature of such schemes, the duty-cycle patterns of neighboring nodes can require a sender to be synchronized with a particular receiver so that it is always selected as the next hop forwarder. This effect, which is called as patterned synchronization effect in this paper, leads to unnecessary drain of energy in packet forwarder nodes reducing their lifetime. We therefore propose a new opportunistic forwarding scheme which is able to identify the occurrence of patterned synchronization effect and resolves it by dynamically adjusting duty-cycle ratio of nodes in a self-organized manner. Using extensive simulations, we show that our scheme significantly improves network lifetime and power utilization of the network while maintaining an acceptable packet transmission delay at the senders.

This paper presents design and implementation of a pipelined datapath for real-time face detection using cascades of boosted classifiers. We propose following methods: symmetric image downscaling, classifier sharing, and cascade merging, to achieve the desired processing speed and area efficiency. First, an image pyramid with 16 levels is generated from the input image to simultaneously detect faces with different scales. The downscaled images are then transferred to the first stage of the cascade that is shared between the corresponding image pairs based on the pixel validity of the symmetric image pyramid. The last method exploits the different hit ratios of the cascade stages. We use a tree-structured cascade of classifiers since most of the nonface elements are eliminated during the early stages of the classifier. The use of a synthesis tool confirms that the proposed design reduces resource utilization by one-eighth without accuracy loss, compared to the fully parallelized implementation of the same algorithm. We implemented the proposed hardware architecture on a Xilinx Virtex-5 LX330 FPGA. The indicative throughput is 307 frames/s irrespective of the number of faces in the scene for standard VGA (640 × 480) images with an operating frequency of 125.59 MHz. We may ensure that face detection results are generated at each clock cycle after the initial pipeline delay, using this fully pipelined datapath for tree-structured cascade classifiers.

ABSTRACT Unlike terrestrial sensor networks, underwater sensor networks (UWSNs) have salient features such as a long propagation delay, narrow bandwidth, and high packet loss over links. Hence, path setup‐based routing protocols proposed for terrestrial sensor networks are not applicable because a large latency of the path establishment is observed, and packet delivery is not reliable in UWSNs. Even though routing protocols such as VBF (vector based forwarding) and HHVBF (hop‐by‐hop VBF) were introduced for UWSNs, their performance in terms of reliability deteriorates at high packet loss. In this paper, we therefore propose a directional flooding‐based routing protocol, called DFR, in order to achieve reliable packet delivery. DFR performs a so‐called controlled flooding, where DFR changes the number of nodes which participate in forwarding a packet according to their link quality. When a forwarding node has poor link quality to its neighbor nodes geographically advancing toward the sink, DFR allows more nodes to participate in forwarding the packet. Otherwise, a few nodes are enough to forward the packet reliably. In addition, we identify two types of void problems which can occur during the controlled flooding and introduce their corresponding solutions. Our simulation study using ns‐2 simulator proves that DFR is more suitable for UWSNs, especially when links are prone to packet loss. Copyright © 2011 John Wiley & Sons, Ltd.

Optical flow is a motion field estimation method that has a wide range of applications. In this paper, we present a fully pipelined hardware architecture for high-speed optical flow estimation based on a full-search block matching algorithm. A census transform is applied to the corresponding pixels in the current and previous frame. The similarity between two census vectors within the search area is then computed by measuring the hamming distance. Macro blocks are generated based on the measured hamming distance values and the best match is determined by locating the block that has the smallest sum. The synthesis tool reported that the proposed system is capable of processing 400 standard VGA frames per second.

Recently, multi-sink wireless sensor networks (WSNs) are envisioned to solve the hot spot problem caused by traditional single-sink WSNs. Routing protocols should be able to distribute network traffic evenly to multiple sinks to prolong network lifetime and they should be scalable. Gradient-based routing protocols are known to be suitable for the networks, where each node maintains its gradient representing the direction toward a neighbor node to reach a sink. In particular, existing protocols allow a sensor node to construct its gradient using the cumulative traffic load of a path for load-balancing. However, they have a critical drawback that a sensor node cannot efficiently avoid using the path with the most overloaded node. Hence, this paper introduces a new gradient-based routing protocol for load-balancing (GLOBAL) with a new gradient model to maximize network lifetime. In GLOBAL, the least-loaded path which also avoids the most overloaded sensor node is selected for forwarding. Through ns-2 simulation, we verify that GLOBAL achieves better performance than the shortest path routing protocol and load-aware gradient-based routing one.

Underwater wireless sensor networks (UWSNs) are prone to message loss due to their poor link condition with high bit error rate (BER). Retransmission-based mechanisms such as ARQ can be used for the reliability, however, they are not suitable for UWSNs due to the inherent long propagation delay of UWSNs. Therefore, a forward error correction (FEC) mechanism might be an alternative in order to guarantee the reliable message delivery. In this paper, we therefore propose a fast and reliable transport (called FRT) protocol for UWSNs. The FRT protocol takes advantage of the FEC mechanism and utilizes link quality toward neighbors so as to provide reliable and fast delivery of packets. Using ns-2 simulator, we verify that our FRT protocol achieves less latency than an existing protocol without an increase of transmission overhead.

This paper presents an FPGA-based parallel hardware architecture for real-time face detection. An image pyramid with twenty depth levels is generated using the input image. For these scaled-down images, a local binary pattern transform and feature evaluation are performed in parallel by using the proposed block RAM-based window processing architecture. By sharing the feature look-up tables between two corresponding scaled-down images, we can reduce the use of routing resources by half. For prototyping and evaluation purposes, the hardware architecture was integrated into a Virtex-5 FPGA. The experimental result shows around 300 frames per second speed performance for processing standard VGA (640times480times8) images. In addition, the throughput of the implementation can be adjusted in proportion to the frame rate of the camera, by synchronizing each individual module with the pixel sampling clock.

Skin color is widely used in many applications because of its merit in human-machine interactions. However, detecting skin color requires repetitive operations on all pixels in the image, similar to other vision-based applications. Since the per-pixel processing is difficult to perform efficiently in conventional computers, many real-time image processing applications have problems with performance. In this paper, we propose FPGA implementation of a real-time skin color detection system. Among the various skin color detection methods, we chose a parametric skin distribution modeling method based on a Gaussian mixture, due to its acceptable training amount and skin detection performance. In addition, a mean-based surface flattening method was also proposed and implemented to improve the detection performance. The proposed method flattens the surface of objects in the scene by replacing the pixel value with the mean of its similar neighborhoods to remove the color noise. After this flattening process, the pixel values of the analogous adjacent pixels are located within a narrow range and are easily segmented to a different region. To consider the inherent parallelism of local image processing, all these functions are implemented within the FPGA to meet the demands of real-time performance.

In delay tolerant networks (DTNs) with frequent network partitioning, routing packets is a challenge, because the successful establishment of an end-to-end path between source and destination nodes is not guaranteed. Typical routing protocols for DTNs depend on data replications over multiple paths for reliable data delivery. Since they invoke a lot of replicated packets, our previous work introduced a Vector Routing protocol utilizing the vector of node movement in order to reduce such replications. In this paper, we extend the previous work and propose a more efficient Vector Routing protocol to exploit some features of history-based routing protocols where packets are forwarded to nodes with higher probability to meet their destination nodes. Unlike the existing protocols where the number of nodal encounters is used to calculate the probability of forwarding, our proposed protocol allows a node to maintain the vector information of other nodes that it has met and share this information with other nodes. Using the history information, a node predicts the location area of each packet's destination node in order to perform more accurate forwarding. The ns-2 simulations using random waypoint and Manhattan mobility models verify that our protocol achieves better performance in terms of less amount of traffic incurred without loss of packet delivery ratio.

Abstract In ad hoc networks, the spatial reuse property limits the number of packets which can be spatially transmitted over a path. In standard Transmission Control Protocol (TCP), however, a TCP sender keeps transmitting packets without taking into account this property. This causes heavy contention for the wireless channel, resulting in the performance degradation of TCP flows. Hence, two techniques have been proposed independently in order to reduce the contention. First, a TCP sender utilizes a congestion window limit (CWL), by considering the spatial reuse property. This prevents the TCP sender from transmitting more than CWL number of packets at one time. Second, a delayed ack (DA) strategy is exploited in order to mitigate the contention between the TCP ACK and DATA packets. Recently, although TCP‐DAA (Dynamic Adaptive Acknowledgment) attempts to utilize a CWL‐based DA strategy, TCP‐DAA overlooks a dynamic correlation between these two techniques. This paper, therefore, reveals the dynamic correlation and also proposes a protocol which not only reduces the frequency of the TCP ACK transmissions but also determines a CWL value dynamically, according to network conditions. Simulation studies show that our protocol performs the best in various scenarios, as compared to TCP‐DAA and standard TCP (such as TCP‐NewReno). Copyright © 2009 John Wiley & Sons, Ltd.

In wireless ad hoc networks, utilizing multiple interfaces as well as multiple channels at MAC layer is known as one of key techniques to improve network capacity. Since the existing multi-channel MAC protocols like MMAC and SSCH assume that a single interface is used, a protocol such as HMCP was proposed in order to make the best use of multiple interfaces. However, a fixed specific channel is assigned to an interface in HMCP. In addition, HMCP requires one interface to accomplish an exclusive purpose of either transmission or reception. To address these shortcomings, we propose a multi-channel multi-interface MAC (MM-MAC) protocol, which dynamically assigns channels to each interface of the nodes and reserves the transmission times of nodes on these selected channels. The scheduled transmission of packets also enables our MM-MAC protocol to reduce the number of channel collisions and the overhead of exchanging control messages. Through extensive study using the ns-2 simulator, we observe that our MM-MAC protocol performs better than the existing protocols such as MMAC and HMCP.

Recently, much research work has paid attention to delay tolerant networks (DTNs), which are networks with a frequent occurrences of network partitioning. Since the successful establishment of an end-to-end path between source and destination nodes is not guaranteed in these networks, routing is a challenging issue. In typical routing protocols for DTNs such as epidemic routing, they depend on data replication techniques over multiple paths for reliable data delivery. However, they invoke a large number of duplicated packets in the network. This paper therefore proposes an efficient routing (called vector routing) by utilizing the vector of node movements. In vector routing, the direction and velocity of nodes are calculated from the location information of nodes, and then nodes efficiently decide which nodes should take replicated packets as well as the number of packets to replicate. Using ns-2 simulation with two different mobility models, namely random waypoint and manhattan mobility models, we verify that vector routing performs better than epidemic routing in terms of less amount of traffic incurred without loss of packet delivery ratio.

IP auto-configuration in mobile ad hoc networks has attracted much attention. Efficient DAD (duplicate address detection) techniques should be devised to provide each node with its unique address in the network. Generally, DAD schemes can be categorized into two classes: (a) active DAD and (b) passive DAD. In this paper, we focus on passive DAD schemes over on-demand ad-hoc routing protocols such as AODV and DYMO. In order to improve the accuracy of detecting address conflicts, we propose several schemes using additional information including sequence, location, or neighbor information

In MANETs, ACK loss can affect TCP communication performance, since end-to-end reliable delivery of data packets is guaranteed by using ACK packets sent from a TCP receiver. This ACK loss may force a TCP sender to produce unnecessary retransmissions (called spurious retransmissions). These spurious retransmissions can also degrade the performance of TCP flows in the network because they will create more channel contention and interferences. These issues have been overlooked and hence require attention. We, therefore, propose new solutions to reduce such spurious retransmissions. Our scheme allows a TCP sender to be correctly informed of the sequence number that the TCP receiver expects to receive (i.e., ACK sequence). This ACK sequence is piggybacked onto routing control packets during a new route discovery process invoked by the TCP receiver. Using the ACK sequence tossed by the routing layer, the TCP sender can reduce spurious retransmissions. We evaluated and compared our scheme with TCP-Reno via simulation on ns-2. Simulation results showed that our proposed solutions outperformed general TCP (for both static and dynamic networks cases under multiple cross TCP traffic) in terms of aggregate throughput and number of spurious retransmissions.

Multiple paths have been used to provide load- balancing of traffic and resilience to node failures in mobile ad-hoc networks or sensor networks. In addition, the energy expenditure can be distributed among nodes over the multiple paths. However, most existing multi-path routing protocols re- quire their complex procedures such as local repair or periodic reconstruction of the multiple paths, during which much loss of packet is even experienced. In this paper, we therefore propose an energy-efficient data gathering technique using multiple paths, which is more suitable for highly populated wireless sensor net- works in terms of having more simple and resilient features than other existing techniques. By simulation using ns-2 simulator, we prove that it achieves two goals satisfactorily: energy-saving and resilience to node failure.

A MANET (mobile ad hoc network) requires an efficient network-wide broadcasting service in route discovery as well as in many applications. Recently, many broadcasting protocols have been proposed to avoid the well-known broadcast-storm problem, or to provide reliability over wireless broadcast media. However, since it is assumed that all nodes in the network have the same radio ranges (i.e. homogeneous radio ranges), the protocols do not work in a realistic environment where all nodes have different fixed radio ranges or they use dynamic radio ranges intentionally to save their energy in the network (i.e. heterogeneous radio ranges). We therefore propose an efficient stateless broadcasting technique to support the heterogeneous radio ranges. An additional area covered through a node's rebroadcasting is calculated by considering different radio ranges of nodes. Then, nodes with larger coverage area will rebroadcast earlier than other nodes with smaller area. When determining rebroadcasting schedule, the DFD (dynamic forwarding delay) concept introduced in DDB (dynamic delayed broadcasting) is exploited in our protocol. Extensive simulation using ns-2 proves that our scheme is more suitable for a realistic environment.

Abstract A great deal of research has been directed toward reducing energy consumption at the 802.11‐based MAC layer for mobile ad hoc networks (MANETs). There exists an approach that uses maximum power during the RTS‐CTS exchange and computes the required amount of power in order for DATA and ACK messages to reach the receiver and sender. It, however, does not consider the existence of an interference range (IR), or often called a carrier sensing zone. In order to address the problem, an existing approach forces nodes, located within a sender's carrier sensing zone, to defer their transmission trials in order to avoid collisions at the sender. It, however, does not consider possible collisions at the receiver, which can result in frequent retransmissions and hence greater energy consumption. Therefore, we propose an efficient protocol called Fragmentation‐based Power Control MAC (F‐PCM), which utilizes the fragmentation mechanism of the IEEE 802.11 MAC protocol so as to avoid collisions at senders as well as those at receivers. Through extensive simulations, our F‐PCM has performed better in terms of providing a higher throughput and incurring lower energy consumption. This is particularly advantageous in a dense mobile network environment where collisions are more severe. Copyright © 2006 John Wiley & Sons, Ltd.

To join an IP network and communicate with others, a node needs to be configured either manually by an administrator or automatically through a DHCP server. However, the former method is impractical for large networks, while the latter is infeasible in the case of a mobile ad hoc network due to the mobility of the nodes. This paper introduces two distributed IP address auto-configuration mechanisms for mobile ad hoc networks, namely (a) RADA (random address allocation) and (b) LiA (linear allocation). RADA is based on random IP address selection, while LiA linearly assigns new addresses by utilizing the current maximum IP address value. We have also introduced an improved version of LiA, known as LiACR (linear allocation with collision resolution), which reduces control overhead. Then, we discuss extensions of these mechanisms capable of handling network partitioning and merging. Performance evaluations of RADA, LiA and LiACR were conducted through simulation. The results related to address allocation time and control overhead are presented and compared.

As interests in MANET (mobile ad hoc networks) have increased, routing protocols suitable for MANET have been standardized specially in IETF (Internet Engineering Task Force). In addition to routing protocols, efforts to adopt the TCP stack as a reliable end-to-end transport protocol with some modifications appropriate for MANETs have been made for the smooth integration with the fixed Internet. Unlike previous approaches to improve TCP performance in MANETs, we modified the fast retransmission technique on the standard TCP. Basically, the fast retransmission technique utilizes the third duplicate ACK to trigger retransmission. However, if TCP is served by on-demand reactive ad-hoc routing protocols like DSR (dynamic source routing), AODV (ad hoc on-demand distance vector) and ABR (associativity-based routing) which can be considered as kinds of connection-oriented routing protocols until a route breakage occurs after a route acquisition, we can improve TCP performance by advancing the retransmission time using the first duplicate ACK triggered retransmission technique, without waiting for the third duplicate ACK packet. We show the performance improvement with this approach by using GIoMoSim simulator with AODV used as its underlying routing protocol, one of standardized on-demand reactive routing protocols in IETF.

Many protocols used in mobile ad hoc networks rely on the broadcasting capability, especially when performing a route discovery process. However, an efficient broadcasting protocol should be devised to reduce the unnecessary redundant rebroadcasting at some nodes (redundancy) as well as to increase the coverage area as much as possible (reachability). A few approaches have been developed in the literature. We propose a bounding algorithm, which is known to be an efficient candidate to accommodate the two goals, that is to increase reachability while limiting redundancy.

Untethered nodes in mobile ad hoc networks strongly depend on the efficient use of their batteries. In this paper, we propose a new metric, the drain rate, to forecast the lifetime of nodes according to current traffic conditions. This metric is combined with the value of the remaining battery capacity to determine which nodes can be part of an active route. We describe new route selection mechanisms for MANET routing protocols, which we call the minimum drain rate (MDR) and the conditional minimum drain rate (CMDR). MDR extends nodal battery life and the duration of paths, while CMDR also minimizes the total transmission energy consumed per packet. Using the ns-2 simulator and the dynamic source routing (DSR) protocol, we compare MDR and CMDR against prior proposals for energy-aware routing and show that using the drain rate for energy-aware route selection offers superior performance results. Methods keywords are system design and simulations.

Mobile ad hoc networks' (MANETs) inherent power limitation makes power-awareness a critical requirement for MANET protocols. We propose a new routing metric, the drain rate, which predicts the lifetime of a node as a function of current traffic conditions. We describe the minimum drain rate (MDR) mechanism which uses a combination of the drain rate with remaining battery capacity to establish routes. MDR can be employed by any existing MANET routing protocol to achieve a dual goal: extend both nodal battery life and connection lifetime. Using the ns-2 simulator and the dynamic source routing (DSR) protocol, we compared MDR to the minimum total transmission power routing (MTPR) scheme and the min-max battery cost routing (MM-BCR) scheme and proved that MDR is the best approach to achieve the dual goal.

Existing routing protocols for mobile ad hoc networks assume that all nodes have the same radio transmission range. In other words, the mobile ad hoc network has symmetric links, which means that two neighboring nodes A and B are within the transmission range of one another. However, since nodes consume battery power independently according to their computing and communication load, there exist asymmetric links, which means that node A is within node B's transmission range, but not vice versa. In this paper, two approaches are presented to support routing in the existence of asymmetric links: GAHA (GPS-based Hop-by-hop Acknowledgment) and GAPA (GPS-based Passive Acknowledgment) schemes. Both GAHA and GAPA can be applied to any routing protocols by utilizing GPS (Global Positioning System) location information. Simulation results reveal that both GAHA and GAPA protocols cope well in the presence of asymmetric wireless links and nodes' mobility.

The existing routing protocols in mobile ad hoc networks assume that all nodes have the same transmission range. In other words, the mobile ad hoc network has only symmetric links. However, since nodes consume battery power independently according to their computing and communication load, there exist asymmetric links, which means that node A is within node B's transmission range, but not vice versa. This paper present two protocols that accommodate asymmetric links: link-level and end-to-end Approaches. The link-level approach can be applied to any routing protocols by utilizing GPS(Global Positioning System) location information of nodes at link level. On the other hand, the end-to-end approach does not need GPS devices and employs dual paths between source and destination. Simulation results reveal that these protocols cope well with ad hoc network having asymmetric wireless links under the presence of mobility.

Reliable data transmission over a wireless multi-hop network, called the ad hoc network, has proven to be non-trivial. The TCP (transmission control protocol), a widely used end-to-end reliable transport protocol in a wired network, is not entirely suitable when applied to a wireless ad hoc network due to TCP's congestion control schemes. In particular, the TCP at the source considers the network as congested when detecting packet losses or timeouts. However, in a wireless ad hoc network when a route disconnection occurs because of node movement, the network mistakes this as a congestion. Therefore, the conventional TCP congestion control mechanism cannot be applied, because a route disconnection must be handled differently from a network congestion. We propose a new mechanism that improves the TCP performance in a wireless ad hoc network where each node can buffer packets during a route disconnection and reestablishment. Additionally, we incorporate new measures to deal with the reliable transmission of important control messages. Our simulation results further confirmed these advantages.

The existing routing protocols in mobile ad hoc networks assume that all the nodes have the same transmission range. In other words, the mobile ad hoc network has only symmetric links, which means that two neighboring nodes A and B are within the transmission range of one another. However, since nodes consume battery power independently according to their computing and communication load, there exist asymmetric links, which means that node A is within node B's transmission range, but not vice versa. In this paper, to cope with asymmetric links, a new routing protocol based on source-initiated on-demand routing (called RODA-routing protocol with dual paths to support asymmetric links) is presented. In order to perform route maintenance in an efficient manner, dual paths-forward (source to receiver) and backward (receiver to source) paths-are used to send and receive data packets. In addition, a timer mechanism is introduced to deal with simultaneous and consecutive broken links on both forward and backward paths. Simulation results reveal that RODA copes well with an ad hoc network having asymmetric wireless links under the presence of mobility.