Search

For the realization of an Internet of Things (IoT) with high densities of devices it is necessary that wireless communication protocols are developed that offer 1) low energy consumption; 2) simplicity of encoding and decoding; 3) an asynchronous mode of communication; and 4) an effective but simple method to deal with interference between transmissions. This paper presents the implementation, experimentation, and analysis of a protocol on the MAC sublayer that encodes information in terms of silent intervals between pulses. Based on the representation of patterns of sparse pulses, this encoding has the potential for extremely low power consumption at the transmitter side. It also results in only few conflicts between messages that are broadcast on the same band overlapped in time, while no synchronization between transmitters and receivers is necessary. The protocol is demonstrated experimentally on the 315 MHz band with 100 senders and one receiver configured in a Star topology. Theoretical analysis confirms that the probability of conflicts between messages is low, even if the number of devices increases to the order of ten thousand. This protocol facilitates the implementation of IoT devices that are restricted in terms of hardware and energy resources.

This letter reports on an ultra-fast and accurate channel assignment approach using a coherent Ising machine (CIM) for dense wireless LAN (WLAN) systems. CIM is a laser-based high-speed hardware approach for solving Ising problems using optoelectronic features. We formulate the total-throughput maximization problem for large-scale centralized WLAN systems as a combinatorial optimization problem. Subsequently, the problem is converted into an Ising problem to be optimized via a CIM. Simulation results reveal that the CIM can improve the total throughput performance and give results in a very short time (millisecond order).

Dynamic channel selection is among the most important wireless communication elements in dynamically changing electromagnetic environments wherein a user can experience improved communication quality by choosing a better channel. Multi-armed bandit (MAB) algorithms are a promising approach by which the difficult tradeoff between exploration to search for better a channel and exploitation to experience enhanced communication quality is resolved. Ultrafast solution of MAB problems has been demonstrated by utilizing chaotically oscillating time series generated by semiconductor lasers. In this study, we experimentally demonstrate a MAB algorithm incorporating laser chaos time series in a wireless local area network (WLAN). Autonomous and adaptive dynamic channel selection is successfully demonstrated in an IEEE802.11a-based, four-channel WLAN. Although the laser chaos time series is arranged prior to the WLAN experiments, the results confirm the usefulness of ultrafast chaotic sequences for real wireless applications. In addition, we numerically examine the underlining adaptation mechanism of the significantly simplified MAB algorithm implemented in the present study compared with the previously reported chaos-based decision makers. This study provides a first step toward the application of ultrafast chaotic lasers for future high-performance wireless communication networks.

We propose a simple channel-allocation method based on tug-of-war (TOW) dynamics, combined with the time scheduling based on nonlinear oscillator synchronization to efficiently use of the space (channel) and time resources in wireless communications. This study demonstrates that synchronization groups, where each node selects a different channel, are non-uniformly distributed in phase space such that every distance between groups is larger than the area of influence. New type of self-organized spatiotemporal patterns can be formed for resource allocation according to channel rewards., 5 pages, 2 figures

Non-orthogonal multiple access (NOMA) can multiplex different users on a single channel by applying superposition coding at the transmitter and Successive Interference Cancellation (SIC) at the receiver. It is considered to be an important technology for next generation wireless communications to achieve high throughput and large-scale connectivity. To take advantage of the effectiveness of NOMA, in addition to pairing users using criteria based on distance and power allocation, a key question is how to optimally allocate resources such as channels to users. Previous studies have discussed user pairing schemes and power allocation algorithms for NOMA systems, but the key issue is how to allocate the channels. Coherent Ising Machine (CIM) is an Ising computer that can solve NP-hard combinatorial optimization problems at high speed. CIM can run optimization algorithms based on mutually coupled neural networks and can be applied to a variety of optimization problems. In this paper, we proposed a CIM-based optimization method for user pairing and channel allocation in NOMA systems. Simulation results show that the CIM-based optimization can achieve superior performance of the NOMA system.

Channel bonding (CB) is one of the most important techniques to optimize the spectrum efficiency in wireless local area networks (WLANs). IEEE 802.11ac as the most widely used standard of WLANs extends the number of basic channels of CB to 8, which can support the maximum channel widths of 160MHz. In order to avoid collisions, improve throughput while reducing communication latency, we propose an ultra-high-speed dynamic CB method using laser decision based MAB algorithm in IEEE 802.11ac network. Our proposed method can make the CB decision within the order of GHz. Besides, most existing works studying the CB protocol are based on analytical model or simulations. Since analytical model or simulations cannot accurately predict the network performance in practical communication scenarios, we evaluate our proposed algorithm by experiments in this paper. Experimental results show that our proposed method can make better CB decisions than other MAB algorithms termed as $\varepsilon$-greedy policy and UCB1-tuned.

It is necessary to develop an efficient channel selection method with low power consumption to achieve high communication quality for distributed massive IoT system. To this end, Ma et al. [1] proposed an autonomous distributed channel selection method based on the Tug-of-War (ToW) dynamics. The ToW-based method can achieve equivalent performance to UCB1-tuned [2], [3] with low computational complexity and power consumption, which is recognized as a best practice technique for solving multi-armed bandit (MAB) problems. However, Ref. [1] only considered fixed IoT devices with simplex communication.

For the realization of an Internet of Things (IoT) with high densities of nodes it is necessary that wireless communication protocols are developed that offer (1) low energy consumption, (2) simplicity of encoding and decoding, (3) an asynchronous mode of communication, and (4) an effective but simple method to deal with interference between transmissions. This paper presents the implementation, experimentation, and analysis of a protocol on the MAC sublayer that is based on the encoding of information by the silent intervals between pulses. This encoding allows for few conflicts between messages that are broadcast on the same band overlapped in time. This is demonstrated experimentally, while no adverse effect is detected on transmissions from broadcasts of a different group of nodes using the same 315 MHz band but with a different code length as coding parameter. Theoretical analysis confirms that the probability of conflicts between messages is low, even if the number of nodes increases to the order of ten thousand. This protocol facilitates the implementation of IoT nodes that are restricted in terms of hardware and energy resources.

We propose a novel database management framework for data delivery services based on Store-Carry-Forward (SCF) techniques. The platform we present consists of heterogeneous wireless opportunistic networks of long-range narrowband and short-range broadband communications. We introduce a blockchain-based method by which to verify the record of delivery events on a decentralized network. A new consensus mechanism named proof-of-forwarding (PoF) is proposed to substitute the function of previously proposed proof-of-work (PoW) methods, while significantly improving computational complexities of block generation. Specifically, in our proposal a block is generated exclusively when data delivery agents perform node-to-node direct communication using a short-range high-speed wireless standard to deliver data. We additionally propose a digital signature overlay to prevent malicious nodes from producing fake transactions without any effort to carry data content to recipients. Simulation results show that our blockchain-based framework robustly manages data delivery records, where 97% of Distributed Denial-of-Service (DDoS) attacks can be prevented even when half of the entire nodes are assumed to be malicious.

The number of IoT devices has dramatically increased. Numerous IoT devices generate enormous traffic, which causes network congestions and packet losses. To deal with network congestions, we have proposed a channel selection algorithm based on reinforcement learning for IoT devices. We have implemented the algorithm on an IoT device with limited function and confirmed autonomous appropriate channel selection in a local area where devices are crowded. In this paper, we conduct evaluation experiments in real environment where devices are deployed distributedly and there are various IoT systems such as Sigfox, LoRaWAN competing for frequency bands. Our experimental results show that the proposed algorithm based on reinforcement learning improves packet delivery rate (frame success rates) and fairness of the network. Furthermore, we apply the machine learning based IoT devices to IoT sensor network for building monitoring. We deploy 30 sensor devices in a 4-floor building and confirm data collection with avoiding congestions and interference by proposed devices.

The number of IoT devices may dramatically increase in the near future. Numerous IoT devices may generate enormous traffic, which causes network congestions and packet losses. To manage network congestions, Ma et al. have proposed a channel selection algorithm based machine learning for IoT devices. They modeled channel selection as Multi-Armed Bandit problem and have designed a algorithm based on Tug-of-War dynamics to solve this problem. Furthermore, they confirmed dynamic channel selection in a local area where devices are crowded. In this paper, we conduct evaluation experimentation in real environment where devices are coexisting with other IoT systems, Sigfox and LoRaWAN. Our experimental results using our implemented systems show that each IoT node selects appropriate channel by the proposed algorithm based on reinforcement learning and the packet delivery rate (frame success rates) and fairness among the sensor nodes can be improved by the proposed scheme.

We are finding the location of damage to timber and wooden houses. Two years ago, we succeeded in classifying the damage location with 90% accuracy in a wooden brace house. Last year, we conducted an experiment on a model house in Oita Prefecture and improved the classification rate by preprocessing data. Therefore, we conducted experiments to further improve the classification rate and practical application. The vibration data of the model house in Oita Prefecture was collected using multiple 3-axis acceleration sensors equipped with wireless communication means and monitored at Katsushika Campus, Tokyo University of Science, about 969 km away. By classifying the waveform data by CNN, we succeeded in classifying the damage location and degree of damage with a maximum accuracy of 86.0%.

Cognitive radio technologies have been developed to utilize the limited spectrum resources and optimize the behavior of the radio system, along with the variable radio status, by intelligent learning. The optimization can be achieved through clarifying the relationship among certain variables and target performance indices. However, today's wireless systems are becoming increasingly complex due to the advances of wireless technologies; this increases the number of variables of the optimization problem, which makes the relations among the variables and system performance to be formulated with difficulty. Recent advance in the field of machine learning technologies can help us overcome such difficulties by adopting for cognitive radio systems. This paper proposes a cross-layer modeling of wireless system using machine learning and optimization method based on cognitive cycle. The experimental evaluation is shown by applying the method to the IEEE 802.11 wireless local area network.

In order to enable the benefits of device-to-device (D2D) communications underlay cellular networks, two fundamental problems need to be addressed: 1) how to control the interference to guarantee the quality-of-service (QoS) of cellular user and 2) how to allocate transmit powers among D2D pairs to maximize the data rate. In this letter, we propose a distributed resource allocation to resolve these problems. Specifically, the interference from D2D transmissions to cellular users is coordinated using a pricing scheme, while a D2D pair competes with other pairs to efficiently reuse the available spectrum. We model this competition using a non-cooperative power control game and propose a distributed update rule to reach the Nash equilibrium. Simulation is provided to evaluate the performance of the proposed algorithm.

We propose the development of a new measurement, assessment, and notification system of building states. This system will utilize the sensing and communication capabilities of the IoT concept. When it detects an emergency or any structural instability within the building, the system will transmit information about the current situation and provide predictions regarding potential danger.

Distributing power packets in a network is an efficient technique for supplying power to distributed loads. In the case that the amount of power is limited in a closed network system, the consensus algorithm can be useful for sharing the limited power resource in the system. For example, we consider a power network consisting of one power source and N(≫1) loads. We assume that each load has a buffer that can store the power for driving the load. It is important to efficiently share the limited power from the single source. One possible way for sharing power is as follows. First, all loads are divided into several clusters. Next, the power source sends power packets to each cluster by responding to the feedback information on how much power is required by the loads in the cluster. If the power demand of one load exceeds the power supply from the source, the deficient power is accommodated by the consensus algorithm among the buffers within the cluster. We discuss by numerical simulations how to cluster the loads in terms of the balance between the cost for distributing power from the source and the performance of loads.

Interference management is one of the most critical issues in underlaying device-to-device (D2D) communication due to the coexistence of D2D pairs and cellular users that operate under the same spectrum. In this paper, we provide the interference management algorithm to maximize the performance of the D2D communication while satisfying the quality-of-service requirements of the cellular communications in both uplink and downlink phases. The proposed algorithm includes: 1) the admission control and power allocation to ensure that the interference from D2D communication does not affect to the cellular communications and 2) the shared channel assignment to maximize the total throughput of the D2D communication. We prove that our proposed algorithm can achieve at least half of the performance of optimal algorithm. The simulation results validate the feasibility, convergence, and optimality of our algorithm: it cannot only closely approximate the optimal throughput of D2D communication but also outperform existing algorithms.

Reinforcement learning involves decision-making in dynamic and uncertain environments and constitutes a crucial element of artificial intelligence. In our previous work, we experimentally demonstrated that the ultrafast chaotic oscillatory dynamics of lasers can be used to efficiently solve the two-armed bandit problem, which requires decision-making concerning a class of difficult trade-offs called the exploration–exploitation dilemma. However, only two selections were employed in that research; hence, the scalability of the laser-chaos-based reinforcement learning should be clarified. In this study, we demonstrated a scalable, pipelined principle of resolving the multi-armed bandit problem by introducing time-division multiplexing of chaotically oscillated ultrafast time series. The experimental demonstrations in which bandit problems with up to 64 arms were successfully solved are presented where laser chaos time series significantly outperforms quasiperiodic signals, computer-generated pseudorandom numbers, and coloured noise. Detailed analyses are also provided that include performance comparisons among laser chaos signals generated in different physical conditions, which coincide with the diffusivity inherent in the time series. This study paves the way for ultrafast reinforcement learning by taking advantage of the ultrahigh bandwidths of light wave and practical enabling technologies.

In this paper, to improve the throughput of Deviceto-device (D2D) communication, we allow D2D to reuse multiple channels of cellular networks. However, since this can cause interference between D2D and cellular network, we use the concept of price payment for reusing the channels to achieve decision making between D2D and cellular network. Here, we propose the Stackelberg game approach to allocate the transmission power of D2D over multi-channel and maximize the price utility of cellular network as well as guarantee the QoS of cellular network. To further improve the utility of D2D, we propose the merge of Stackelberg game with overlapping coalition formation game to select the multi-channel with small interference to D2D. The results show the improvement in utility of D2D using the latter proposed method.

SUMMARY We propose a natural synchronization scheme for wireless uncoupled devices, without any signal exchange among them. Our proposed scheme only uses natural environmental fluctuations, such as the temperature or humidity of the air, the environmental sounds, and so on, for the synchronization of the uncoupled devices. This proposed synchronization is realized based on the noise-induced synchronization phenomenon, uncoupled nonlinear oscillators synchronize with each other only by adding identical common noises to each of them. Based on the theory of this phenomenon, the oscillators can also be synchronized by noise sequences, which are not perfectly identical signals. Since the environmental natural fluctuations collected at neighboring locations are similar to each other and cross-correlation becomes high, our proposed scheme enabling synchronization only by natural environmental fluctuations can be realized. As an application of this proposed synchronization, we introduce wireless sensor networks, for which synchronization is important for reducing power consumption by intermittent data transmission. We collect environmental fluctuations using the wireless sensor network devices. Our results show that the wireless sensor network devices can be synchronized only by the independently collected natural signals, such as temperature and humidity, at each wireless sensor device.

We propose a new cognitive radio network architecture using the IP multimedia subsystem (IMS) functionality. We implement the cognitive radio network entities standardized in IEEE 1900.4 on the IMS that exchanges RAN and terminal context information between the networks and the terminals to make optimum and immediate reconfiguration decisions. In our proposed architecture, RAN context information is obtained from cellular networks which are directly connected to the IMS. The presence management functions of the IMS are applied to exchange those information in a “push” manner, which enables immediate notification of changes in wireless environment. We evaluate the performance of the proposed context information exchange method, by comparing with the cases that adequate and immediate RAN context information is not available. The evaluation results show that the proposed framework gives 10-30% superior performance than the conventional cognitive radio networks.

It is important to optimize aggregation schemes for heterogeneous wireless networks for maximizing communication throughput utilizing any available radio access networks. In the heterogeneous networks, differences of the quality of service (QoS), such as throughput, delay and packet loss rate, of the networks makes difficult to maximize the aggregation throughput. In this paper, we firstly analyze influences of such differences in QoS to the aggregation throughput, and show that it is possible to improve the throughput by adjusting the parameters of an aggregation system. Since manual parameter optimization is difficult and takes much time, we propose an autonomous parameter tuning scheme using a machine learning algorithm for the heterogeneous wireless network aggregation. We implement the proposed scheme on a heterogeneous cognitive radio network system. The results on our experimental network with network emulators show that the proposed scheme can improve the aggregation throughput better than the conventional schemes. We also evaluate the performance using public wireless network services, such as HSDPA, WiMAX and W-CDMA, and verify that the proposed scheme can improve the aggregation throughput by iterating the learning cycle even for the public wireless networks. Our experimental results show that the proposed scheme achieves twice better aggregation throughput than the conventional schemes.

The tabu search was implemented on a neural network with chaotic neuro-dynamics. This chaotic exponential tabu search shows great performance in solving quadratic assignment problems (QAPs). To exploit inherent parallel processing abilities of analog hardware systems, a synchronous updating scheme, where all the neurons in the network are updated at the same time, was proposed. However, several neurons may fire simultaneously with the synchronous updating. As a result, we cannot determine only one candidate for the 2-opt exchange from the many fired neurons. To solve this problem, several neuron selection methods, which select one specific neuron among the fired neurons, were proposed. These neuron selection methods improved the performance of the synchronous updating scheme. In this paper, we analyze the dynamics of the chaotic neural network with the neuron selection methods by means of the spatial and temporal mutual information. Through the analyses, the network solution search dynamics of the exponential chaotic tabu search with different neuron selection methods are evaluated.

We propose an autonomous access point selection algorithm for user-centric radio resource usage optimization in distributed wireless networks. We introduce the optimization algorithm based on the mutually connected neural network, which minimizes a given objective function by distributed update of each neuron. In order to improve the quality of services for each user, we apply such an algorithm to optimization of the balance of the available throughput among the users with keeping higher average throughput per user. The mutually connected neural network to minimize the objective function is realized by calculating the connection weights and the thresholds from the coefficients of the energy function and the target objective function. By computer simulations, we show that the proposed algorithm improves the available throughput for each user in large-scale wireless networks. Furthermore, we implement the proposed algorithm on an experimental wireless network, and verify that each user terminal selects a most appropriate access point to optimize the total radio resource usage based on the state of neurons distributively updated at each user terminal.

The mobile communications market has grown rapidly over the past ten years, but the market could reach saturation in the foreseeable future. More flexible mobile networks that can meet various user demands and create new market openings are required for further growth. Heterogeneous networks are more suitable than homogeneous networks for meeting a wide variety of user demands. There are two types of heterogeneous networks: a closed type, where network resources are deployed and operated by communication carriers, and an open type, where network resources can be deployed not only by existing operators, but also by companies, universities, and so on. It will be easy for newcomers to enter mobile businesses in an open heterogeneous mobile network so many innovative services are likely to be provided through cooperation between various companies or organizations. This article proposes a revised architecture for TISPAN-NGN, which corresponds to heterogeneous networks and open mobile markets, and presents a new business model.

We propose a neurodynamical approach to a large-scale optimization problem in Cognitive Wireless Clouds, in which a huge number of mobile terminals with multiple different air interfaces autonomously utilize the most appropriate infrastructure wireless networks, by sensing available wireless networks, selecting the most appropriate one, and reconfiguring themselves with seamless handover to the target networks. To deal with such a cognitive radio network, game theory has been applied in order to analyze the stability of the dynamical systems consisting of the mobile terminals' distributed behaviors, but it is not a tool for globally optimizing the state of the network. As a natural optimization dynamical system model suitable for large-scale complex systems, we introduce the neural network dynamics which converges to an optimal state since its property is to continually decrease its energy function. In this paper, we apply such neurodynamics to the optimization problem of radio access technology selection. We compose a neural network that solves the problem, and we show that it is possible to improve total average throughput simply by using distributed and autonomous neuron updates on the terminal side.

金沢大学医薬保健研究域保健学系, We have developed and reported super-high resolution liquid crystal displays (SHR-LCDs) using a new resolution enhancement technology of the independent sub-pixel driving (ISD) that utilizes three sub-pixels contained in each pixel element. This technology realizes the three-times resolution enhancement of monochrome LCDs. A 15 mega-pixel (MP) SHR-LCD out of a 5MP LCD and a 9MP SHR-LCD out of a 3MP LCD, for example, are realized by this technology, which improves the depiction ability of detailed image shapes such as micro-calcifications of a mammography. Furthermore, the ISD technology brings not only resolution enhancement but also noise reduction effect by the high-frequency data sampling in displaying the clinical images. In this study, we have investigated the clinical efficacy of the SHR-LCDs by means of phantom observation studies and blind observer comparison studies using clinical mammography images performed by radiologists. We used a conventional 5MP LCD for a comparison of a 15MP SHR-LCD and a 9MP SHR-LCD to evaluate their efficacy. From the results of the studies, it was indicated that the SHR-LCDs using the ISD technology had the excellent ability to display the high-resolution digital mammography images. © 2008 Springer-Verlag Berlin Heidelberg.

A context-based adaptive communication system is introduced for use in heterogeneous networks. Context includes the user's presence, location, available network interfaces, network availability, network priority, communication status, terminal features, and installed applications. An experimental system was developed to clarify the feasibility of using context information to flexibly control networks and applications. The system operates on a seamless networking platform we developed for heterogeneous networks. By using contexts, the system can inform the caller and callee of applications they can access, which are available through the network before communication occurs. Changes in contexts can switch an on-going application to another during actual communication. These functions provide unprecedented styles of communication. A business scenario for a seamless networking provider is also presented.

For future generation mobile networks, we expect that the mobile devices like PDAs, note PCs or any VoIP-enabled communicators will have the feature of being always switched on, ready for service, constantly reachable by the wireless Internet. In addition to high access speed, attractive real-time contents or other expected spectacular features of the future wireless Internet environment, the mobile terminals has to be very much energy-aware to enable literal untethered movement of the user. Mechanisms for network activities like maintaining location information and wireless system discovery, which require regular network access, should be energy-efficient and resource-efficient in general. Cellular systems employ the notion of passive connectivity to reduce the power consumption of idle mobile hosts. In IP based Multi-service User Terminal (MUT) that may have multiple wireless interfaces for receiving various classes of services from the network, there should be an efficient addressing of the energy consumption issue. To devise an energy-efficient scheme for simultaneous or single operation of the wireless interfaces attached to such terminals we should have comprehensive understanding of the power consumption of the devices/modules in various operational states. This paper investigates the power consumption pattern or behavior of some selected wireless interfaces that are good candidates for being part of the future of the multi-service user terminals. We propose a simple model for predicting energy consumption in a terminal attributed to the wireless network interfaces. We measured the actual consumption pattern to estimate the parameters of the model.

Cognitive radio technology improves efficiency of radio resource usage and quality of services in congested mobile networks, by reconfiguring the wireless connection settings according to the optimum decisions, which are made based on the collected context information. This paper focuses on the optimization algorithms for the decision-making to optimize radio resources usage in heterogeneous cognitive wireless networks. For the networks with centralized management, we can apply the algorithms to obtain rigorous solutions, because all of the network context information can be collected at the central manager. In order to avoid exponential increase of computational complexity in large-scale wireless networks, we model the target optimization problem as a minimum cost flow problem, and make it possible to solve the problem in polynomial time. In the networks with distributed management, the central network manager is not necessary, but the algorithms have to be run in a distributed manner. In order to optimize the defined objective function, we introduce the distributed energy minimization of the Hopfield-Tank neural network. We derive the decision-making rule for each terminal to optimize the entire network. We show validity of the proposed algorithms by numerical simulations.

We show that the wireless sensor networks can be synchronized by natural environmental signals, such as temperature, humidity and so on. The proposed synchronization scheme is based on the noise-induced phase synchronization theory; the phases of the periodical limit cycle orbits of nonlinear oscillators synchronize with each other by adding a common noise signal to the oscillators. Based on this theory, we synchronize the clocks of the wireless sensor nodes by tuning them according to the phase of their nonlinear oscillator to which natural environmental signals are added as noise. In this paper, first we analyze cross-correlation of the natural environmental signals measured by ZigBee wireless sensor nodes, which are arranged outdoor at about 20m intervals, and show that the cross- correlation among the signals sensed at different sensor nodes are around 0.8 or higher. According to this result, we analyze possibility of the noise- induced phase synchronization with changing cross- correlation between the additive noises, and clarify that the nonlinear oscillators can be synchronized in the cases that the cross-correlation becomes around 0.8 or higher. Finally, we investigate feasibility of the noise-induced phase synchronization by actual data of natural environmental signals sensed at each sensor node and show it is possible to synchronize wireless sensor nodes without any interactions or communications among them.

This paper shows autonomous optimization techniques for distributed cognitive radio architecture. Based on the context information exchange mechanism standardized in IEEE1900.4, we assume that the terminals can obtain various types of context information of radio access networks (RANs) and they can make optimal decisions by themselves using the shared information. In our research, we apply the neural network based algorithms, the learning algorithms and the distributed optimization algorithms, to realize autonomous optimization of the cognitive networks. We apply the proposed schemes to the heterogeneous type cognitive radio, whose radio resource usage can be optimized by appropriate RAN selection. First, we apply learning algorithms to appropriate parameter selection for each RAN to maximize the aggregated throughput using multiple RANs. We show that our developed autonomous learning system can autonomously improve the throughput. We also apply a neural network based optimization technique to optimize radio resource usage of the entire network. We show that the proposed scheme can distributively optimize the entire network as an autonomous cognitive dynamic wireless system.

By increase of various radio access network (RAN) services, available spectrum resources for mobile communications get decrease, and efficient use of the radio resource is becoming a very important issue. In order to optimize the radio resource usage and maxmize the throughput and quality of service (QoS), the link aggregation technologies to utilize multiple different available RANs have been studied. However, in such heterogeneous wireless networks, it is difficult to improve the throughput by their aggregation because of the differences among the QoSs of the different RANs. In this paper, we propose an autonomous parameter optimization scheme using a machine learning algorithm, which maximize the throughput of the heterogeneous RAN aggregation system. We evaluate the performance of the proposed scheme implemented on a cognitive wireless network system called Cognitive Wireless Cloud (CWC) system, connected to real wireless network services, such as HSDPA, WiMAX and W-CDMA. Our experimental results of the proposed system show that the aggregation throughput can be improved with increase of the training samples, which are collected autonomously.

We propose a context-aware guide application, which provides appropriate information selected by a machine learning algorithm according to the preference and the situation of each user. We have designed and implemented the proposed system using the off-the-shelf mobile phones with a built-in GPS module. The machine learning algorithm enables our system to select an appropriate spot based on the user’s real-time context such as preference, location, weather, time, etc. As a machine learning algorithm, we use the support vector machine (SVM) to decide the appropriate information for the users. In order to realize high generalization performance, we introduce the principal component analysis (PCA) to generate the input data for the SVM learning. By our experiments in real environments, it is shown that the proposed system works correctly and the correctness of recommendation can be improved by introducing the PCA.

This paper investigates effectiveness of distributed radio access network (RAN) selection schemes to optimize radio resource usage, for the realistic heterogeneous wireless networks, in which the distribution of the base stations are non-uniform and the non-cognitive legacy terminals also connect to the same RANs. Our proposed algorithm is based on mutually-connected and higher-order neural networks, which can directly optimize any objective functions of entire heterogeneous wireless networks, by completely distributed procedure. Our simulation results show that the proposed scheme is effective even for non-uniform heterogeneous wireless network environments, especially for higher-order and complex optimization problems. Such improvements in the proposed method also can be seen also for the cases that the legacy non-cognitive terminals co-exist in the same heterogeneous wireless networks.

In heterogeneous wireless network environment, various radio access networks (RANs) are available for mobile terminals which have multiple wireless network interfaces to connect to each different RAN. However, it is very difficult to improve the throughput by aggregating different RANs, which have different quality of service (QoS) in the throughput, the delay, the packet loss, and so on. In this paper, we propose a traffic allocation control scheme using a machine learning algorithm to improve the link aggregation throughput in such heterogeneous wireless networks. We introduce the support vector machine, which has an advantage on generalization to have good performance for unknown input patterns. By our experiments using a cognitive wireless network system, called Cognitive Wireless Clouds, it is clarified that our proposed scheme improves the throughput higher than the conventional methods based on the throughput of each RAN.

This paper proposes a method that stabilizes chaotic phenomena, which appears in a user-centric cognitive radio networks by the chaos control theory. In the cognitive radio networks, spectrum resources are shared among the users in order to optimize efficiency of radio resource usage. However, depending on the algorithm of the spectrum sharing procedure, undesirable chaotic oscillation may occur in the network. For such chaotic nonlinear systems, the chaos control theory has been proposed and applied to a wide variety of nonlinear systems in order to stabilize such systems. In this paper, the delay feedback method is introduced as one of the chaos control algorithms, and applied to control of the chaotic phenomena in the cognitive radio networks. We show that our method can stabilize the chaotic oscillations in the usercentric cognitive radio networks. By comparisons with traditional parameter tuning methods, we confirm that our method is more efficient and faster to stabilize the cognitive radio systems.

柵工は植生復旧を促すことを目的とする治山工種である。これには木杭が大量に消費されている。そこで,木杭の実際の耐用年数を求めるため,富山県内の柵工を中心とする44箇所の施工地を延べ111回調査した。調査は抜取りと現地調査の二法で行った。抜取り調査では採取木杭の玉切り試料から被害度と縦圧縮強さの関係を,現地調査では木杭15本の地際部平均被害度と折損率および経過年数との関係を調べた。被害度の評価は森林総合研究所の6段階評価法(表－1)に準拠して行った。得られた結果は以下のとおりである。1)調査した全ての木杭が無処理のまま用いられていた。樹種はスギとカラマツが多く,両者で約85%に達した。寸法は平均で頂部直径約11cm,長さおよそ1.2m,設置地上部長さは約40cmであった。2)縦圧縮強さは試片の被害度が大きくなるにつれて直線的に低下し,被害度5(破壊)で0となった。3)地際部平均被害度と折損率の関係プロットから,折損率は10%付近からの急激な上昇が認められた。この折損率を柵工の耐用限界とすると,被害度は約3.4と計算され,スギ,カラマツ木杭の推定耐用年数はそれぞれ約7.4, 8.7年となった。

In this paper, we apply distributed optimization dynamics of the mutually connected neural networks to RAN selection in heterogeneous type cognitive wireless networks. We evaluate the performance of the proposed approach by implementing it on an experimental heterogeneous wireless network system called Cognitive Wireless Cloud, which supports vertical handover between different radio access networks and various information exchange defined in IEEE1900.4. Our neural algorithm implemented on such an experimental wireless network optimizes objective function without any centralized computation. As the objective functions, we introduce two types of problems, load balancing and QoS satisfaction rate optimization, and compare the performance of the proposed method with those of other distributed RAN selection algorithms on the real wireless system. The experimental results using such a system show that the proposed algorithm exhibits the best performance. Since our algorithm based on the neural network dynamics directly optimizes the objective functions defined for radio resource usage optimization of the entire wireless network by distributed computation on each terminal, its performance becomes better than other algorithm which is based on the improvement of each terminal's QoS.