We present M3-SLU, a new multimodal large language model (MLLM) benchmark for evaluating multi-speaker, multi-turn spoken language understanding. While recent models show strong performance in speech and text comprehension, they still struggle with speaker-attributed reasoning, the ability to understand who said what and when in natural conversations. M3-SLU is built from four open corpora (CHiME-6, MELD, MultiDialog, and AMI) and comprises over 12,000 validated instances with paired audio, transcripts, and metadata. It includes two tasks: (1) Speaker-Attributed Question Answering and (2) Speaker Attribution via Utterance Matching. We provide baseline results for both cascaded pipelines and end-to-end MLLMs, evaluated using an LLM-as-Judge and accuracy metrics. Results show that while models can capture what was said, they often fail to identify who said it, revealing a key gap in speaker-aware dialogue understanding. M3-SLU offers as a challenging benchmark to advance research in speaker-aware multimodal understanding.

Image Manipulation Detection and Localization (IMDL) aims to identify tampered images and their altered regions. Existing RGB-centered approaches often overemphasize RGB information while overlooking complementary insights from noise-view modalities. This reliance on RGB limits their ability to detect subtle manipulation traces. To overcome these challenges, we propose Multi-Noise-View Fusion (MNVFusion), a framework that balances the contributions of RGB and noise-view modalities using a multi-branch encoder structure. MNVFusion incorporates the Multi-Branch Channel Mixing Module (MB-CMM), enabling efficient channel-wise fusion to integrate diverse modality features. Additionally, we introduce Fixed GeM, a training-free image-level detection module that enhances overall efficiency through fixed operations on localization maps. Experiments on six benchmark datasets show that MNVFusion delivers state-of-the-art performance in both detection and localization tasks.

Recent advances in computer vision have introduced generalized image segmentation models applicable across various domains. However, camouflaged object detection (COD) remains a particularly challenging task that requires dedicated approaches, owing to the minimal visual distinction between objects and their backgrounds. The degree of camouflage is influenced by three critical aspects—color, texture, and edge—requiring methodologies that address these simultaneously. Efforts to detect camouflaged objects have continually focused on these aspects. In this study, we propose the Tri-Aspects Network (TANet) for COD, designed to overcome the limitations of existing approaches that primarily focus on a single aspect. TANet emphasizes differences in color, texture, and edge to detect camouflaged objects. It consists of an ensemble of two independent networks that learn from the color differences extracted through color conversion and the textural features extracted using Bayar convolution filters. Each independent network enhances high-level features extracted from the input image through the Context Enhancement Block (CEB) and maximizes the difference between the background and camouflaged objects during reconstruction with the prediction mask using the Multi-scale Edge Refinement Block (MERB). The results from these two networks are then ensembled. Additionally, by using an erosion kernel to ensure that the prediction mask’s edge closely matches the ground truth edge, more fine-grained predictions can be achieved. TANet proposes a novel COD network that shows outstanding results compared to existing models in three key evaluation metrics (S-measure, E-measure, and weighted F-score), demonstrating its contribution to the field.

Telemonitoring is the use of electronic devices such as smartphones to remotely monitor patients. It provides great convenience and enables timely medical decisions. To facilitate the decision making for each patient, a model is needed to translate the data collected by the patient’s smartphone into a predicted score for his/her disease severity. To train a robust predictive model, semi-supervised learning (SSL) provides a viable approach by integrating both labeled and unlabeled samples to leverage all the available data from each patient. There are two challenging issues that need to be simultaneously addressed in using SSL for this problem: (1) feature selection from high-dimensional noisy telemonitoring data; and (2) instance selection from many, possibly redundant unlabeled samples. We propose a novel SSL model allowing for simultaneous feature and instance selection, namely the S2SSL model. We present a real-data application of telemonitoring for patients with Parkinson’s Disease using their smartphone-collected activity data such as tapping and speaking. A total of 382 features were extracted from the activity data of each patient. 74 labeled and 563 unlabeled instances from 37 patients were used to train S2SSL. The trained model achieved a high accuracy of 0.828 correlation between the true and predicted disease severity scores on a validation dataset. Note to Practitioners—Telemonitoring is an emerging health care platform enabled by smartphones and wearables. Because it allows for health data to be collected anytime and anywhere, patients can be frequently monitored and medical decisions can be made more timely and effectively. This paper addresses the data science challenges in leveraging the telemonitoring platform to benefit patient care. Specifically, we propose a new model, S2SSL, to tackle these challenges and provide better robustness, accuracy, and efficiency. This paper may be interesting to health care practitioners seeking advanced analytics capabilities to model and integrate the data collected through telemonitoring devices, with ultimate purposes of improving the decision in treating each patient and increasing patient access to specialized care.

The instrumentation and control (I&C) system of a nuclear power plant (NPP) employs a cybersecurity program regulated by the government. Through regulation, the government requires the implementation of security controls in order for a system to be developed and operated. Accordingly, the licensee of an NPP works to comply with this requirement, beginning in the development phase. The compliance-driven approach is efficient when the government supervises NPPs, but it is inefficient when a licensee constructs them. The security controls described in regulatory guidance do not consider system characteristics. In other words, the development organization spends a considerable amount of time excluding unnecessary control items and preparing the evidence to justify their exclusion. In addition, security systems can vary according to the developer’s level of security knowledge, leading to differences in levels of security between systems. This paper proposes a method for a developer to select the appropriate security controls when preparing the security requirements during the early development phase; it is designed to ensure the system’s security and reduce the cost of excluding unnecessary security controls. We have formalized the representation of attack patterns and security control patterns and identified the relationships between these patterns. We conducted a case study applying RG 5.71 in the Plant Protection System (PPS) to confirm the validity of the proposed method.

Modern deep learning frameworks rely heavily on GPUs to accelerate the computation. However, the security implication of GPU device memory exploitation on deep learning frameworks has been largely neglected. In this paper, we argue that GPU device memory manipulation is a novel attack vector against deep learning systems. We present a novel attack method leveraging the attack vector, which makes deep learning predictions no longer different from random guessing by degrading the accuracy of the predictions. To the best of our knowledge, we are the first to show a practical attack that directly exploits deep learning frameworks through GPU memory manipulation. We confirmed that our attack works on three popular deep learning frameworks, TensorFlow, CNTK, and Caffe, running on CUDA. Finally, we propose potential defense mechanisms against our attack, and discuss concerns of GPU memory safety.

Abstract Deep neural networks (DNNs) provide superior performance on machine learning tasks such as image recognition, speech recognition, pattern analysis, and intrusion detection. However, an adversarial example, created by adding a little noise to an original sample, can cause misclassification by a DNN. This is a serious threat to the DNN because the added noise is not detected by the human eye. For example, if an attacker modifies a right-turn sign so that it misleads to the left, autonomous vehicles with the DNN will incorrectly classify the modified sign as pointing to the left, but a person will correctly classify the modified sign as pointing to the right. Studies are under way to defend against such adversarial examples. The existing method of defense against adversarial examples requires an additional process such as changing the classifier or modifying input data. In this paper, we propose a new method for detecting adversarial examples that does not invoke any additional process. The proposed scheme can detect adversarial examples by using a pattern feature of the classification scores of adversarial examples. We used MNIST and CIFAR10 as experimental datasets and Tensorflow as a machine learning library. The experimental results show that the proposed method can detect adversarial examples with success rates: 99.05% and 99.9% for the untargeted and targeted cases in MNIST, respectively, and 94.7% and 95.8% for the untargeted and targeted cases in CIFAR10, respectively.

In the last decade, the dedicated graphics processing unit (GPU) has emerged as an architecture for high-performance computing workloads. Recently, researchers have also focused on the isolation property of a dedicated GPU and suggested GPU-based secure computing environments with several promising applications. However, despite the security analysis conducted by the prior studies, it has been unclear whether a dedicated GPU can be leveraged as a secure processor in the presence of a kernel-privileged attacker. In this paper, we first demonstrate the security of dedicated GPUs through comprehensive studies on context information for GPU execution. The paper shows that a kernel-privileged attacker can manipulate the GPU contexts to redirect memory accesses or execute arbitrary GPU codes on the running GPU kernel. Based on the security analysis, this paper proposes a new on-chip execution model for the dedicated GPU and a novel defense mechanism supporting the security of the on-chip execution. With comprehensive evaluation, the paper assures that the proposed solutions effectively isolate sensitive data in on-chip storages and defend against known attack vectors from a privileged attacker, supporting that the commodity GPUs can be leveraged as a secure processor.

Deep neural networks (DNNs) display good performance in the domains of recognition and prediction, such as on tasks of image recognition, speech recognition, video recognition, and pattern analysis. However, adversarial examples, created by inserting a small amount of noise into the original samples, can be a serious threat because they can cause misclassification by the DNN. Adversarial examples have been studied primarily in the context of images, but their effect in the audio context is now drawing considerable interest as well. For example, by adding a small distortion to an original audio sample, imperceptible to humans, an audio adversarial example can be created that humans hear as error-free but that causes misunderstanding by a machine. Therefore, it is necessary to create a method of defense for resisting audio adversarial examples. In this paper, we propose an acoustic-decoy method for detecting audio adversarial examples. Its key feature is that it adds well-formalized distortions using audio modification that are sufficient to change the classification result of an adversarial example but do not affect the classification result of an original sample. Experimental results show that the proposed scheme can detect adversarial examples by reducing the similarity rate for an adversarial example to 6.21%, 1.27%, and 0.66% using low-pass filtering (with 12 dB roll-off), 8-bit reduction, and audio silence removal techniques, respectively. It can detect an audio adversarial example with a success rate of 97% by performing a comparison with the initial audio sample.

Deep neural networks (DNNs) are widely used for image recognition, speech recognition, and other pattern analysis tasks. Despite the success of DNNs, these systems can be exploited by what is termed adversarial examples. An adversarial example, in which a small distortion is added to the input data, can be designed to be misclassified by the DNN while remaining undetected by humans or other systems. Such adversarial examples have been studied mainly in the image domain. Recently, however, studies on adversarial examples have been expanding into the voice domain. For example, when an adversarial example is applied to enemy wiretapping devices (victim classifiers) in a military environment, the enemy device will misinterpret the intended message. In such scenarios, it is necessary that friendly wiretapping devices (protected classifiers) should not be deceived. Therefore, the selective adversarial example concept can be useful in mixed situations, defined as situations in which there is both a classifier to be protected and a classifier to be attacked. In this paper, we propose a selective audio adversarial example with minimum distortion that will be misclassified as the target phrase by a victim classifier but correctly classified as the original phrase by a protected classifier. To generate such examples, a transformation is carried out to minimize the probability of incorrect classification by the protected classifier and that of correct classification by the victim classifier. We conducted experiments targeting the state-of-the-art DeepSpeech voice recognition model using Mozilla Common Voice datasets and the Tensorflow library. They showed that the proposed method can generate a selective audio adversarial example with a 91.67% attack success rate and 85.67% protected classifier accuracy.

Image synthesis is a novel solution in precision medicine for scenarios where important medical imaging is not otherwise available. The convolutional neural network (CNN) is an ideal model for this task because of its powerful learning capabilities through the large number of layers and trainable parameters. In this research, we propose a new architecture of residual inception encoder-decoder neural network (RIED-Net) to learn the nonlinear mapping between the input images and targeting output images. To evaluate the validity of the proposed approach, it is compared with two models from the literature: synthetic CT deep convolutional neural network (sCT-DCNN) and shallow CNN, using both an institutional mammogram dataset from Mayo Clinic Arizona and a public neuroimaging dataset from the Alzheimer's Disease Neuroimaging Initiative. Experimental results show that the proposed RIED-Net outperforms the two models on both datasets significantly in terms of structural similarity index, mean absolute percent error, and peak signal-to-noise ratio.

Deep neural networks (DNNs) perform well in the fields of image recognition, speech recognition, pattern analysis, and intrusion detection. However, DNNs are vulnerable to adversarial examples that add a small amount of noise to the original samples. These adversarial examples have mainly been studied in the field of images, but their effect on the audio field is currently of great interest. For example, adding small distortion that is difficult to identify by humans to the original sample can create audio adversarial examples that allow humans to hear without errors, but only to misunderstand the machine. Therefore, a defense method against audio adversarial examples is needed because it is a threat in this audio field. In this paper, we propose a method to detect audio adversarial examples. The key point of this method is to add a new low level distortion using audio modification, so that the classification result of the adversarial example changes sensitively. On the other hand, the original sample has little change in the classification result for low level distortion. Using this feature, we propose a method to detect audio adversarial examples. To verify the proposed method, we used the Mozilla Common Voice dataset and the DeepSpeech model as the target model. Based on the experimental results, it was found that the accuracy of the adversarial example decreased to 6.21% at approximately 12 dB. It can detect the audio adversarial example compared to the initial audio sample.

Telemonitoring is the use of electronic devices to remotely monitor patients. Taking the Parkinson's disease (PD) as an example, the use of at-home testing device (AHTD) enables remote, internet-based measurement of PD vocal symptoms. Translating AHTD measurement into a unified PD rating scale (UPDRS) through predictive analytics enables cost-effective, convenient, and close tracking of PD progression. Building a predictive model between AHTD measurement and UPDRS is not straightforward because PD patients are highly heterogeneous, which requires patient-specific models. Learning a patient-specific model faces the challenge of limited data. Transfer learning (TL) tackles this challenge by leveraging other patients' information to make up the data shortage when modeling a target patient. Among different TL methods, the category of parameter transfer methods is more appropriate for the telemonitoring application because it transfers patient-specific model parameters but not patients' data. However, existing parameter transfer methods fall short because not every other patient's information is helpful and blind transfer causes the problem of negative transfer. To tackle this limitation, we propose a positive TL (PTL) method. We provide an in-depth theoretical study on the risk and condition for negative transfer to happen, which further drive the development of novel PTL algorithms that are robust to negative transfer. We apply PTL to predict UPDRS of 42 PD patients using their AHTD vocal measurement. PTL achieves significantly better accuracy compared with single learning and one-model-fits-all approaches.

Deep neural networks (DNNs) exhibit excellent performance in machine learning tasks such as image recognition, pattern recognition, speech recognition, and intrusion detection. However, the usage of adversarial examples, which are intentionally corrupted by noise, can lead to misclassification. As adversarial examples are serious threats to DNNs, both adversarial attacks and methods of defending against adversarial examples have been continuously studied. Zero-day adversarial examples are created with new test data and are unknown to the classifier; hence, they represent a more significant threat to DNNs. To the best of our knowledge, there are no analytical studies in the literature of zero-day adversarial examples with a focus on attack and defense methods through experiments using several scenarios. Therefore, in this study, zero-day adversarial examples are practically analyzed with an emphasis on attack and defense methods through experiments using various scenarios composed of a fixed target model and an adaptive target model. The Carlini method was used for a state-of-the-art attack, while an adversarial training method was used as a typical defense method. We used the MNIST dataset and analyzed success rates of zero-day adversarial examples, average distortions, and recognition of original samples through several scenarios of fixed and adaptive target models. Experimental results demonstrate that changing the parameters of the target model in real time leads to resistance to adversarial examples in both the fixed and adaptive target models.

An intrusion tolerant system (ITS) is a network security system that is composed of redundant virtual servers that are online only in a short time window, called exposure time. The servers are periodically recovered to their clean state, and any infected servers are refreshed again, so attackers have insufficient time to succeed in breaking into the servers. However, there is a conflicting interest in determining exposure time, short for security and long for performance. In other words, the short exposure time can increase security but requires more servers to run in order to process requests in a timely manner. In this paper, we propose Duo, an ITS incorporated in SDN, which can reduce exposure time without consuming computing resources. In Duo, there are two types of servers: some servers with long exposure time (White server) and others with short exposure time (Gray server). Then, Duo classifies traffic into benign and suspicious with the help of SDN/NFV technology that also allows dynamically forwarding the classified traffic to White and Gray servers, respectively, based on the classification result. By reducing exposure time of a set of servers, Duo can decrease exposure time on average. We have implemented the prototype of Duo and evaluated its performance in a realistic environment.

There are recommendation systems everywhere in our daily life. The collection of personal data of users by a recommender in the system may cause serious privacy issues. In this article, we propose the first privacy-preserving matrix factorization for recommendation using fully homomorphic encryption. Our protocol performs matrix factorization over encrypted users’ rating data and returns encrypted outputs so that the recommendation system learns nothing on rating values and resulting user/item profiles. Furthermore, the protocol provides a privacy-preserving method to optimize the tuning parameters that can be a business benefit for the recommendation service providers. To overcome the performance degradation caused by the use of fully homomorphic encryption, we introduce a novel data structure to perform computations over encrypted vectors, which are essential for matrix factorization, through secure two-party computation in part. Our experiments demonstrate the efficiency of our protocol.

Recommendation systems become popular in our daily life. It is well known that the more the release of users' personal data, the better the quality of recommendation. However, such services raise serious privacy concerns for users. In this paper, focusing on matrix factorization-based recommendation systems, we propose the first privacy-preserving matrix factorization using fully homomorphic encryption. On inputs of encrypted users' ratings, our protocol performs matrix factorization over the encrypted data and returns encrypted outputs so that the recommendation system knows nothing on rating values and resulting user/item profiles. It provides a way to obfuscate the number and list of items a user rated without harming the accuracy of recommendation, and additionally protects recommender's tuning parameters for business benefit and allows the recommender to optimize the parameters for quality of service. To overcome performance degradation caused by the use of fully homomorphic encryption, we introduce a novel data structure to perform computations over encrypted vectors, which are essential operations for matrix factorization, through secure 2-party computation in part. With the data structure, the proposed protocol requires dozens of times less computation cost over those of previous works. Our experiments on a personal computer with 3.4 GHz 6-cores 64 GB RAM show that the proposed protocol runs in 1.5 minutes per iteration. It is more efficient than Nikolaenko et al.'s work proposed in CCS 2013, in which it took about 170 minutes on two servers with 1.9 GHz 16-cores 128 GB RAM.

Last encounter routing (LER) is an excellent routing paradigm that exploits distributed mobility diffusion to achieve both moving object tracking and packet networking services in dynamic mobile networks. From our observations, we discover that LER can be easily extended to a mobile DHT protocol that introduces excellent performance to high-speed mobility environments. This is of particular interest when higher level of mobility and membership dynamics go hand in hand. In our simple but powerful DHT paradigm, a data publish/look-up process consists of a sequence of spatial motion tracking of the rendezvous node that is responsible for the data resource. Thus, we name the protocol MX-DHT (Motion-MiX-DHT). As opposed to existing topology based DHT schemes, MX-DHT does not require additional management of logical or overlying look-up topologies, except for the one-hop encounter records of logical metadata carried by mobile nodes. Therefore, in high-speed mobility and dynamic membership environments, MX-DHT achieves a significant reduction in the communication costs of the publish/look-up and join/leave operations as compared to existing mobile DHT schemes. An extensive set of experiments showed that MX-DHT is a cost-effective solution to providing a content centric networking service in different types of networks with dynamic mobility and membership changes.

In this paper, we propose an efficient multihop broadcast protocol for asynchronous duty-cycled wireless sensor networks (EMBA) where each node independently wakes up according to its own schedule. EMBA adopts two techniques of the forwarder's guidance and the overhearing of broadcast messages and ACKs. A node transmits broadcast messages with guidance to neighbor nodes. The guidance presents how the node forwards the broadcast message to neighbor nodes by using unicast transmissions. This technique significantly reduces redundant transmissions and collisions. The overhearing of broadcast messages and ACKs helps to reduce the number of transmissions, thus it minimizes the active time of nodes. We implement EMBA and conventional protocols of ADB and RIMAC broadcast in ns-2 simulator to compare their performance. The simulation results show that EMBA outperforms ADB and RI-MAC broadcast in both sparse and dense networks. EMBA achieves lower message cost than the conventional protocols and significantly improves the energy efficiency in terms of both duty cycle and energy consumption.

To reduce the number of unnecessary handover in hierarchical macro/femto-cell networks, it is necessary to avoid macro → femto cell handovers of temporary femtocell visitors who stay in the femtocell for a relatively short time. In this paper, we propose a smart handover decision algorithm exploiting future mobility pattern prediction scheme to prevent macro → femto cell handovers of such temporary femtocell visitors. Our simulation result shows that the proposed algorithm effectively reduces the number of unnecessary handovers.

Two major goals of queue management are flow fairness and queue-length stability However, most prior works dealt with these goals independently. In this paper, we show that both goals can be effectively achieved at the same time. We propose a novel scheme that realizes flow fairness and queue-length stability. In the proposed scheme, high-bandwidth flows are identified via a multilevel caching technique. Then, we calculate the base drop probability for resolving congestion with a stable queue, and apply it to individual flows differently depending on their sending rates. Via extensive simulations, we show that the proposed scheme effectively realizes flow fairness between unresponsive and TCP flows, and among heterogeneous TCP flows, while maintaining a stable queue.

In the near future, various multicast based services will be provided over wireless mesh networks. For secure multicast services, various tree based group key management schemes have been introduced until now. Traditional tree based approaches mainly focus on reducing the number of rekeying messages transmitted by the key distribution center. However, they do not consider the network bandwidth used for transmitting each rekeying message. We propose a bandwidth efficient key tree management scheme for dynamic wireless mesh networks where membership changes occur frequently. Simulation results show that our scheme effectively reduces the bandwidth consumption used for rekeying compared to existing key tree schemes.

IEEE 802.16e uses EAP-based authentication and key management for link layer security. Due to the lack of ability to support mobility, however, EAP-based key management becomes a principal impediment to the achievement of an efficient and secure handover in IEEE 802.16e mobile WiMAX networks. In this paper, an overview of the EAP-based handover procedures of the latest IEEE 802.16e standard is given and their security flaws are analyzed. Possible solutions for secure handover in IEEE 802.16e networks are also proposed in this paper. The proposed handover protocol guarantees a backward/forward secrecy while gives little burden over the previous handover protocols.

We propose an optimal modulation and coding scheme (MCS) selection criterion for maximizing user throughput in cellular networks. The proposed criterion adopts both the Chase combining and incremental redundancy based hybrid automatic repeat request (HARQ) mechanisms and it selects an MCS level that maximizes the expected throughput which is estimated by considering both the number of transmissions and successful decoding probability in HARQ operation. We also prove that the conventional MCS selection rule is not optimized with mathematical analysis. Through link-level and system-level simulations, we show that the proposed MCS selection criterion yields higher average cell throughput than the conventional MCS selection schemes for slowly varying channels.

An improved base-phis expansion method is proposed, in which the bit-length of coefficients is shorter and the number of coefficients is smaller than in Kobayashi's expansion method. The proposed method meshes well with efficient multi-exponentiation algorithms. In addition, two efficient algorithms based on the proposed expansion method, named phis-wNAF and phis-SJSF, are presented which significantly reduce the computational effort involved in online precomputation by using the property of Frobenius endomorphism. The proposed algorithms noticeably accelerate computation of a scalar multiplication on Koblitz curves over optimal extension fields (OEFs). In particular, for OEFs where the characteristic is close to 32 bits or 64 bits, the required number of additions is reduced up to 50% in comparison with Kobayashi's base-phis scalar multiplication algorithm. Finally, a method that significantly reduces the memory usage of the precomputation table at the expense of slightly more computation is presented.

Abstract As research in ad hoc networking has advanced, the importance of security increases more and more. But it suffers from the restriction that ad hoc networks do not have the established infrastructure. In this paper we propose a security‐enhanced model with ID‐based cryptosystem which removes the necessity for any infrastructure and provides sound authentication. Also we provide a more secure and concrete routing protocol with aggregate signature. For confidentiality, we also provide a key exchange protocol. And our model not only distributes the role of key generation center in ID‐based cryptosystem to the nodes with threshold cryptography, but also increases availability by providing a secret re‐sharing protocol. Therefore, our model is very appropriate for ad hoc networks. Copyright © 2007 John Wiley & Sons, Ltd.

Recently, a number of fair queueing algorithms for wireless networks have been proposed. They, however, need perfect channel prediction before transmission and rarely consider a medium access control (MAC) algorithm. In the wireless world, the link level retransmission scheme is popularly used in the MAC layer for recovering channel errors. Therefore, we propose a new wireless fair queueing algorithm that works well with link level retransmission and does not require channel prediction. Through simulation, we showed that our algorithm guarantees throughput and fairness. Also, we found that our algorithm achieves flow separation and compensation.

Next generation wireless communication systems are expected to satisfy quality of service (QoS) requirements by preventing call droppings which are caused by user mobility. Call admission control (CAC) is becoming more important in guaranteeing the QoS of calls. Since existing CAC schemes estimate future requirements and reserve bandwidth to guarantee a certain level of QoS, the bandwidth estimation method is very important in CAC schemes. Most conventional schemes use the movement information of individual mobile terminals (MTs) to estimate the bandwidth requirements. While these schemes estimate the future bandwidth usage relatively well, it requires a high computational overhead and lots of memory space to keep and update the movement information of each mobile terminal, and therefore is hard to implement. We propose a call admission control scheme, which estimates the future handoffs using the handoff histories between BSs and the MT conformity. The proposed scheme achieves a low dropping ratio and high bandwidth utilization keeping a low computational overhead. The performance of the proposed scheme is evaluated through simulations.

We focus on delivering radio engineers a practical design of macro-micro code division multiple access (CDMA) cellular overlays. First, we review our algorithmic approach to jointly deploy macrocells and microcells over today's big urban areas having spatially nonuniform traffic distributions. Next, we identify several further issues related to the optimal design of macro-micro cellular overlays and enhance the cell-deploying algorithm to reflect these issues. The numerical results by extensive event-driven simulations show that the resulting macro-micro cellular overlays successfully cope with the existing conditions of today's big urban areas, such as spatial and temporal traffic distributions and user mobility characteristics. Finally, we discuss the practical guidelines for designing macro-micro cellular overlays in the existing big urban areas.

Because the Internet traffic, that will be the major traffic of broadband integrated services digital networks, is bursty when cells are being switched within the multistage switching network, it has a higher possibility that multiple cells arriving simultaneously at a switching element through different incoming links may have to be forwarded along the same outgoing link. We propose a high-performance large-scale ATM switch dealing with such link contention problem. It is a new unbuffered augmented Banyan network using fully adaptive self-routing control: the deflection self-routing Banyan network. To utilize all the links of the network as alternate paths, we employ the deflection-routing algorithm in each switching element, such that cells failing to get selected for the intended link are sent along different links, in the hope that they later return, or detour the contended link and continue their journey to the destination. Cells are never dropped within the switching network, whereas the switch has no multiple cell buffers. The proposed routing is as simple as that of the generic Banyan network, and all the switch elements (SEs) have a uniform structure. To design the proposed network and its self-routing, we use the topological properties that all the SEs of the Banyan network are arranged in a regular pattern topologically. We formulate and prove these properties through an algebraic formalism. We also ran a performance analysis to provide quantitative comparison against the Banyan network and the replicated Banyan networks. As a result, we show that the new network has a far better performance and scalability than the other networks.

B-ISDNs are required to support a variety of services such as audio, data, and video, so that the guarantee of quality-of-service (QoS) has become an increasingly important problem. An effective fair scheduling algorithm permits high-speed switches to divide link bandwidth fairly among competing connections. Together with the connection admission control, it can guarantee the QoS of connections. We propose a novel fair scheduling algorithm, called "virtual-time-based round robin (VTRR)". Our scheme maps the priorities of packets into classes and provides service to the first non-empty class in each round. Also, it uses an estimation method of the virtual time necessary to this service discipline. To find the first non-empty class, the VTRR adopts a priority queueing system of O(loglog c) which decreases the number of instructions which need to be carried out in one packet transmission time segment. These policies help the VTRR implementation in software, which presents flexibility for upgrades. Our analysis has demonstrated that the VTRR provides bounded unfairness and its performance is close to that of weighted fair queuing. Therefore, the VTRR has a good performance as well as simplicity, so that it is suitable for high-speed B-ISDN.

We study fault-tolerant multicasting in multistage interconnection networks (MINs) for constructing large-scale multicomputers. In addition to point-to-point routing among processor nodes, efficient multicasting is critical to the performance of multicomputers. This paper presents a new approach to provide fault-tolerant multicasting, which employs the restricted header encoding schemes. The proposed approach is based on a recursive scheme in order to send a multicast packet to the desired destinations detouring faulty element(s). In the proposed fault-tolerant multicasting, a multicast packet is routed to its own destinations in only two passes through the MIN having a number of faulty elements by exploiting its nonblocking property.

A multistage interconnection network is a suitable class of interconnection architecture for constructing large-scale multicomputers. Broadcast and multicast communication are fundamental in supporting collective communication operations such as reduction and barrier synchronization. In this paper, we propose a new multicast technique in wormhole-switched bidirectional multistage Banyan networks for constructing large-scale multicomputers. To efficiently support broadcast and multicast with simple additional hardware without deadlock, we propose a two-phase multicast algorithm which takes only two transmissions to perform a broadcast and a multicast to an arbitrary number of desired destinations. We encode a header as a cube and adopt the most upper input link first scheme with periodic priority rotation as arbitration mechanism on contented output links. We coalesce the desired destination addresses into multiple number of cubes. And then, we evaluate the performance of the proposed algorithm by simulation. The proposed two-phase multicast algorithm makes a significant improvement in terms of latency. It is noticeable that the two-phase algorithm keeps broadcast latency as efficient as the multicast latency of fanout 2m where m is the minimum integer satisfying 2/sup m//spl ges//spl radic/(N) (N is a network size).

The submesh allocation problem is to recognize and locate a free submesh that can accommodate a request for a submesh of a specified size. In this paper, we propose a new best-fit submesh allocation strategy for mesh-connected multiprocessor systems. The proposed strategy maintains and uses a free submesh list for an efficient allocation. For an allocation request, the strategy selects the best-fit submesh which causes the least amount of potential processor fragmentation. As many large free submeshes as possible are preserved for later allocations. For this purpose, we introduce a novel function quantifying the degree of potential fragmentation of submeshes. The proposed strategy has the capability of recognizing a complete submesh. We also propose an allocation strategy for faulty meshes which can maintain and allocate virtual submeshes derived from faulty submeshes. Extensive simulation is carried out to compare the proposed strategy with previous strategies. The proposed strategy has the best performance: a 6-50 percent improvement over the previous best strategy.

We study wormhole routed networks and their suitability for real-time traffic in a priority-driven paradigm. A traditional blocking flow control in wormhole routing may lead to a priority inversion in the sense that high priority packets are blocked by low priority packets for unlimited time. This uncontrolled priority inversion causes the frequent deadline missing. This paper therefore proposes a new flow control called throttle and preempt flow control, where high priority packets can preempt network resources held by low priority packets, if necessary. As a result, this flow control does not cause priority inversion. Our simulations show that the throttle and preempt flow control dramatically reduces deadline miss ratio without extra virtual channels. It is also observed that the throttle and preempt flow control offers shorter delay for non-real-time traffic than existing real-time flow control does.

The sequential prefetching scheme is a simple hardware controlled scheme, which exploits the sequentiality of memory accesses to predict which blocks will be read in the near future. We analyze the relationship between the sequentiality of application programs and the effectiveness of sequential prefetching on shared-memory multiprocessors. Also, we propose a simple hardware scheme which selects the prefetching degree on each miss by adding a small table (PDS: Prefetching Degree Selector) to the sequential prefetching scheme. This scheme could prefetch consecutive blocks aggressively for applications with high sequentiality and conservatively for applications with low sequentiality.

In this paper, we study multicasting in the self-routing multistage interconnection networks (MINs) for asynchronous transfer mode (ATM) switch architectures. Many B-ISDN applications require multicast connections in addition to conventional point-to-point connections. This paper presents a novel approach to support multicast connection, on the basis of a restricted address encoding scheme which constructs a short fixed-size multicast header and a recursive scheme that recycles a multicast packet one or more times through the network to send it to the desired destinations. The proposed two-phase multicast algorithm provides deadlock-free multiple multicast connections in MIN-based ATM switches. The emphasis is on analyzing the performance of an unbuffered MIN-based switch using the multicast algorithm in terms of network throughput. The proposed algorithm can be easily applied to buffered MIN-based ATM switches.

Recently, a new routing algorithm called inside-out routing algorithm was proposed for routing an arbitrary permutation in the omega-based 2log/sub 2/ N stage networks. This paper discusses the problems of the inside-out routing algorithm and shows that the suggested condition for proper routing in the omega-omega network is insufficient. An extended necessary and sufficient condition for proper routing in the omega-omega network is also suggested. However, it is unknown if any permutation can be successfully routed by a heuristic algorithm which follows the condition. Thus, the rearrangeability of the omega-omega network still remains an open problem.

We discuss the multicast communication in the self-routing multistage interconnection network (MIN) for constructing the internal architecture of asynchronous transfer mode (ATM) switches. Many applications of ATM switches require multicast communications in addition to conventional point-to-point communications. This paper presents a novel approach to supporting multicast communication, on the basis of the recursive scheme that recycles a multicast packet one or more times through the network to reach at desired destinations. We also propose cost-effective multicast algorithms providing deadlock-freedom in MIN-based ATM switches. The proposed algorithms require a small and fixed number of recycling passes and a reasonable number of links used. The proposed algorithms can be easily applicable to buffered MIN-based ATM switches.

We consider a LAN-based multimedia information system like a Video On Demand (VOD) system that supports the retrieval of continuous media like motion video and sound. In the system, the server transmits the streams of continuous media on a shared communication channel while the continuity of multiple streams should be preserved. Several scheduling algorithms have been studied to guarantee the temporal constraints of time-critical messages, but there has been no study to schedule periodic transmission requests with variable bit rates (VBR), which results from the compression algorithms for motion video and sound. We suggest real-time scheduling algorithms that one is static and the other is dynamic, and an admission control algorithm to guarantee the delivery of continuous media. The characteristics of our algorithms are that it is nonpreemptive to save the overheads of preemption, and the static scheduling algorithm is proved to be optimal. It is shown through simulations that the performance of our dynamic scheduling algorithm is better than that of nonpreemptive Earliest Deadline First (EDF) algorithm, especially, under the assumption of variable bit rates.

An ATM switch based on multistage interconnection networks with multiple paths can support higher bandwidth than that of buffered networks with single path by passing multiple packets to the same destination simultaneously. These multiple packets are buffered in the output buffer of the destination. The performance of output buffer in the switch with multiple paths is closely dependent on the output traffic distribution, which is the packet arrival rate at each output link destined to a given output port. As the nonuniformity of the output traffic distribution becomes higher, the performance of the output buffer as like delay and packet loss probability gets better. In this paper, we propose a new self-routing switch architecture with multiple paths called Fly network and analyze the performance of the switching network and the output buffer focusing on the output traffic distribution. It is shown that the Fly network supports best throughput and latency of the output buffer by producing a high degree of the nonuniform output traffic distribution.

The knockout switch is a nonblocking, high-performance switch suitable for broadband packet switching. It allows packet losses, but the probability of a packet loss can be kept extremely small in a cost-effective way. The performance of the knockout switch was analyzed under uniform traffic. In this paper, we present a new, more general analytic model of the knockout switch, which enables us to evaluate the knockout switch under nonuniform traffic. The new model also incorporates the effects of a concentrator and a shared buffer on the packet loss probability. Numerical results for nonuniform traffic patterns of interest are presented.>

Multimedia refers to the composition of multiple monomedia which should be synchronized temporally and spatially.Over the past few years, some trials to describe the composition and synchronization have been made in a variety of applications and these descriptions have been used as frameworks in their applications.Although conventional works have succeeded in describing the synchronizations well, we indicate that they do not deal with the interactivity required in Interactive Multimedia Applications(IMA) like coursewares and hypermedia systems.In this paper, we propose a new specification method based on Milner's Calculus of Communicating Systems(CCS) to cope with the interactivity.For showing the effectiveness of the specification, we design and implement a visual programming environment based on the specification mechanism, and propose a simple courseware as a programming example.Our approach has implications that the new specification mechanism can be adapted as a framework in various interactive applications and the visual programming environment acquires the benefits that it can handle user interactions and synchronizations with only visual expressions while additional texts for control commands should be augmented in conventional works.

The authors describe an extended backpropagation algorithm for fully connected recurrent neural networks applied to speech recognition. The extended delta rule is approximated by excluding some of the past activities of the dynamic neurons to reduce computational complexity without performance degradation. In speaker-dependent recognition of a confusable syllable set, the fully recurrent neural network with the approximated backpropagation algorithm showed better performance than the multilayer perceptron and the self-recurrent network with comparable time complexity. In addition, it is found that most self-recurrent connections become excitatory and most mutual recurrent connections become inhibitory.>

The authors propose three indirect star-type networks, the indirect star networks I and II and the star-delta network, and investigate their properties. An indirect star-type network is obtained by unfolding the star graph. The star-delta network is obtained through an unfolding scheme based on the recursive property of the star graph, and has n-1 switching stages. The star-delta network has the advantage of being controlled by the destination tag routing scheme. The indirect star-type networks are to the star graph as the indirect cube-type networks are to the n-cube. The authors analyze the performance of the indirect star-type networks under uniform traffic to investigate their potential as an alternative to the indirect cube-type networks for the future high-performance large multiprocessor systems.>

A multistage interconnection network (MIN) for multiprocessor systems is proposed. The proposed MIN, called the B-network, uses backward links to provide backward paths for the requests blocked at switches or memory due to contentions. The gamma network is known to contain a cube network (specifically, the inverse omega network) as a substructure. The B-network is obtained from the gamma network by preserving the cube structure but reversing the direction of all other links. These backward links are used as alternate paths for requests blocked due to path or memory contentions. The B-network can be controlled by the simple destination tag control algorithm; packets navigating through the B-network, using both regular forward links and backward links, can reach their destinations under the destination tag control. The performance of the B-network is analyzed under the uniform traffic model and compared to various networks of interest. It is shown that the B-network surpasses the performance of the gamma network, the crossbar switch, and single-buffered MINs based on (2*2) switches, while having the same hardware complexity as the gamma network.>

An analytic model and analytic results for the performance of multibuffered packet-switching interconnection networks in multiprocessor systems are presented. The performance of single-buffered delta networks is first modeled using the state transition diagram of a buffer. The model is then extended to account for multiple buffers. The analytic results for multibuffered delta networks are compared to simulation results. The performance of multibuffered data manipulator networks is analyzed to demonstrate the generality of the model.>

A scheme based on the standard multiplier recoding technique that enables the use of (5*5) switches for PM2I networks is presented. The connections between switching stages are based on the modified PM2I functions, and are called plus-minus-2/sup 2i/ (PM22I); hence, these networks are called the PM22I networks. Since the number of switching stages in the PM22I networks is one-half of those in the PM2I networks, with a moderate increase in the switch size from (3*3) to (5*5), these networks provide similar design tradeoffs availability by the (2*2) and (4*4) cube networks.>