2011 Annual Report of Advanced Networked Environment Division, Cybermedia Center, Osaka University, Japan

This web page is the 2011 annual report of Advanced Networked Environment Division, Cybermedia Center, Osaka University (Ubiquitous Network Laboratory, Graduate School of Information Science and Technology, Osaka University / Nakano Laboratory).

Japanese version of this web page is here.

Table of contents

  1. Members
    1. Laboratory staffs
    2. Joint researcher
    3. Students
  2. Achievements
    1. Research on wireless mesh networks
      1. Time slot assignment algorithms considering adaptive modulation and coding in IEEE 802.16j relay networks
      2. Time slot assignment mechanism considering adaptive modulation and coding in IEEE 802.16j relay networks
      3. Time slot assignment method based on estimation of radio wave interference in IEEE 802.16j relay networks
      4. An adaptive link scheduling by graph coloring for wireless mesh networks with partial traffic observation capability
      5. Effective node repositioning method with limitation of the number of repositioned nodes for IEEE 802.16j relay networks
      6. Performance evaluation and estimation of IEEE 802.16j relay networks considering radio wave obstruction by obstacles
    2. Research on object counting and tracking
      1. Counting pedestrians in video sequences based on optical flow clustering
      2. Counting pedestrians using H.264/AVC video stream
      3. Counting pedestrians passing through virtual gate in video sequences based on optical flow clustering
      4. Counting pedestrians with compound-eye binary sensors
    3. Research on overlay network architectures
      1. Reduction of inter-ISP transit cost caused by overlay routing
      2. Proactive recovery from multiple failures utilizing overlay networking technique
      3. Scalable and density-aware measurement strategies for overlay networks
      4. The implication of overlay routing on ISPs' connecting strategies
    4. Research on architecture of transport layer for next-generation high-speed network
      1. Analysis of power consumption in TCP data transmission over a wireless LAN environment
      2. Influence of power saving technique in IEEE 802.11 multi-hop networks on network performance
      3. Improvement of end-to-end bandwidth measurement method for power-saving routers
      4. Measuring available bandwidth of multiple parts on end-to-end network path
      5. Design and evaluation of a data transfer mechanism for video streaming over TCP
      6. Applications for inline network measurement technique to adaptive quality cntrol in video conference system
      7. Improving interactive user experience on thin clients
      8. Performance evaluation of distributed computing environment considering power consumption
  3. Publications
    1. Book chapters
    2. Journal papers
    3. International conference papers
    4. Theses
      1. Ph.D. Theses
      2. Masters' Theses
      3. Bachelors' Theses

1. Members

1.1. Laboratory staffs

Hirotaka Nakano
Hirotaka Nakano
Professor
Go Hasegawa
Go Hasegawa
Associate Professor
Yoshiaki Taniguchi
Yoshiaki Taniguchi
Assistant Professor
Mari Onishi
Mari Onishi
Secretary

1.2. Joint researcher

Masayuki Murata
Masayuki Murata
Osaka University
Graduate School of IST
Professor

1.3. Students

Shoji Kurakake
Shoji Kurakake
Doctoral Course 3rd
Yasuhiro Ogiri
Yasuhiro Ogiri
Doctoral Course 3rd
Masafumi Hashimoto
Masafumi Hashimoto
Doctoral Course 2nd
Kazuhito Matsuda
Kazuhito Matsuda
Doctoral Course 2nd
Xun Shao
Xun Shao
Doctoral Course 2nd
Yuki Ise
Yuuki Ise
Master Course 2nd
Daisuke Kobayashi
Daisuke Kobayashi
Master Course 2nd
Takafumi Shigefuji
Takafumi Shigefuji
Master Course 2nd
Shizuka Fujisawa
Shizuka Fujisawa
Master Course 2nd
Tomoya Kitayama
Tomoya Kitayama
Master Course 1st
Wataru Tourisaka
Wataru Tourisaka
Master Course 1st
Ryosuke Mine
Ryosuke Mine
Master Course 1st
Yusuke Iijima
Yusuke Iijima
Bachelor Course 4th
Shoichi Takagi
Shoichi Takagi
Bachelor Course 4th
Kohei Higo
Kohei Higo
Bachelor Course 4th
Shuto Fujii
Shuto Fujii
Bachelor Course 4th
Miki Mizushima
Miki Mizushima
Bachelor Course 4th

2. Achievements

2.1. Research on wireless mesh networks

2.1.1. Time slot assignment algorithms considering adaptive modulation and coding in IEEE 802.16j relay networks

Time slot assignment algorithms considering adaptive modulation and coding in IEEE 802.16j relay networks

The adaptive modulation and coding in wireless networks increases the amount of the data transmitted by a wireless link while maintaining the transmission quality. However, in IEEE 802.16j wireless multi-hop relay networks, there is no research on the performance evaluation and improvement of a time slot assignment algorithm considering the adaptive modulation and coding. In this report, we propose a time slot assignment method for the IEEE 802.16j multi-hop relay networks in consideration of the signal to noise ratio in the data transmission between relay nodes. The proposed method improves the network performance by adjusting the number of time slots assigned to links according to transmission quality. Through performance evaluations, we confirm that the data transfer throughput is increased by up to 50% by using our method.

[Related paper]

  1. Tomoya Kitayama, Go Hasegawa, Yoshiaki Taniguchi, and Hirotaka Nakano ``Proposal and evaluation of time slot assignment algorithms considering adaptive modulation and coding in IEEE 802.16j relay networks,'' IEICE Technical Report (IN2011-20), vol.111, no.67, pp.13-18, May 2011. (in Japanese)

2.1.2. Time slot assignment mechanism considering adaptive modulation and coding in IEEE 802.16j relay networks

Time slot assignment mechanism considering adaptive modulation and coding in IEEE 802.16j relay networks

A gateway node and relay nodes construct a tree-form network with radio connection and the network enable wide range connection service in IEEE 802.16j wireless multi-hop relay network.A client terminal is connected with a relay node that exists within communication area and can reach backhaul network by connecting with gateway node on multi-hop communication.In IEEE 802.16j networks, Time Division Multiplexing Access (TDMA) that assign time slots to links between nodes is used to avoid ratio wave interference.Because a link can connect only with time slots that the link is assigned, links can avoid ratio wave interference by assigning different time slot to links that ratio wave interference will occur if links communicate at the same time. On the other hand, when the adaptive modulation and coding is used for encoding of wireless links, communication quality of links decreases due to communication of links that assign a same time slot, as a result, transmission rate of links decreases.To make sure of adequate transmission rate handling traffic load of links, it seems valid to increase the number of time slots but it occurs enlargement of frame size at the same time, so transmission quantity of unit time doesn't seem to increase effectively. Because many of existing research related to time slot assignment methods in IEEE 802.16j networks treat the interference relation on the basis of the simple model, they can't consider this relationship between ratio wave interference and transmission rate. Therefore, time slot assignment method that can grasp more precise interference relation and can evaluate relationship between ratio wave interference and transmission rate is required.

In this research, we propose time slot assignment method supposing rayleigh fading environments that can consider interference relation of links assumed more realistic environment.The proposed method assign time slots considering dynamic changes of interference relation of links in rayleigh fading environments.Specifically, on the time slot assignment process, it estimates deterioration of communication quality that links communicate at the same time with the model and increases the number of time slots that the link is assigned if it can't make sure of transmission rate corresponding to the traffic demand. In this way, the proposed method assign time slots so that all links make sure of transmission rate corresponding to the traffic demand. Performance evaluation of the proposed method is execute by the simulation that use frame size which is the total number of time slots to assign to all links as peformance index. Expressly, we quantitatively evaluate relationship between a qualification related to communication quality and frame size. As a result, we reveal that it keeps frame size small while assigning time slots corresponding to the traffic demand to each link by adjusting the qualification related to communication quality.

[Related paper]

  1. Kohei Higo, ``Time slot assignment mechanism considering adaptive modulation and coding in IEEE 802.16j relay networks,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)

2.1.3. Time slot assignment method based on estimation of radio wave interference in IEEE 802.16j relay networks

Time slot assignment method based on estimation of radio wave interference in IEEE 802.16j relay networks

Multi-hop relay network based on IEEE 802.16j standard is that a gateway node and two or more relay nodes build a tree-based network by wireless communications. It is possible to provide wireless broadband environment for wide area. Time Division Multiple Access is used by IEEE 802.16j standard, that is to say communication opportunity called time slot is assigned to link, and link communications only in the assigned time slot. Consequently, it is possible to perform data transmission to avoid radio interference by assigning time slots to the links in the networks to avoid radio interference caused by multiple links to communicate at the same time. In this case, using a lot of time slots in order to avoid interference, the total number of time slots called frame size is increased, the transmission opportunity per unit time of each link is reduced. On the other hand, by assigning the same time slot to multiple links does not interfere with each other to communicate at the same time, it is possible to reduce the frame size. To assign time slots in this way, it is necessary to understand the relationship between the link interference.

Many existing research about Time slot assignment in IEEE802.16j relay networks has been using a simple model called protocol interference model to determine the relationship of link interference. This model has the advantage in their relationships to determine two values depending on the distance between nodes only the presence or absence of interference, that easily handle problems such as time slot assignment. However, in order to make decisions more realistic interference, consider the signal to interference ratio is the ratio of the strength of the received signal and an interference radio is enabled. And also, according to the signal-to-interference ratio, to dynamically change the amount of transmitted in the time slot has been assigned link is effective to improve the network performance. But in the protocol interference model cannot consider them. Therefore, time slot assignment method that assumes a more detailed model of radio interference, and performance evaluation of relay networks based on it is required.

In this research, we proposed a method to determine the order of the link to assign a time slot. It is important in the time slot assignment method that in IEEE 802.16j relay networks assumes the model it is possible to obtain a detailed interference relationship. In the proposed method to determine the order of the links based on transmission quality of each link which is estimated based on the model which was assumed to transmit all link in the network at the same time. By assigning a time slot based on the order was determined by considering the transmission quality of the link, it is possible to reduce the frame size. We evaluate the performance of the proposal method by comparison with time slot assignment method which has been used in protocol interference model of using the indicator of frame size. As a result, by using the proposed method, it was shown that the proposal method can be reduced the frame size of up to 10% compared to comparative method. In addition, we compared with the method repeating time slot assignment based on the order in which it was randomly determined. It was shown that the proposed method gives a slightly larger frame size than the random method and the calculation time is short.

[Related paper]

  1. Shoichi Takagi, ``Time slot assignment method based on estimation of radio wave interference in IEEE 802.16j relay networks,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)

2.1.4. An adaptive link scheduling by graph coloring for wireless mesh networks with partial traffic observation capability

An adaptive link scheduling by graph coloring for wireless mesh networks with partial traffic observation capability

We proposed a method for a link scheduling of a wireless mesh network that adapts to the change of average traffic demands at communication links. This method is based on a weighted graph coloring and we came up with new weight updating procedure to adapt to traffic demands change. We conducted data transmission simulations where traffic demands at randomly selected communication nodes increase to compare the adaptability of the proposed method and conventional methods to traffic demands increment. The upper bound of traffic demands increment where all data were successfully transmitted to destinations was investigated as an indicator of the throughput of the network. The upper bound is dependent on the percentage of nodes that increase traffic demands. When the percentage of nodes that increase traffic demands grows, the upper bounds decrease for the proposed method and an non-adaptive conventional method, however, the ratio of the upper bound of traffic demands increment of the proposed method to that of the non-adaptive conventional method were kept around 200% for all percentages. It is concluded that the proposed method acquired significant adaptivity for traffic demands increment compared to conventional methods. The effect of the restriction of traffic demands observation to limited nodes was also investigated. We examined the upper bounds of traffic demands for networks designed by the proposed method with four different types of traffic load observation restriction. All restricted networks did not show the adaptability when the percentage of the number of observable nodes is low, however as the percentage grows, the ratio of upper bound of traffic demands of each method to that of the non-adaptive method reaches 121% to 174%. Among the restrictions, the restriction to nodes with large traffic demands at initial time showed the superiority for most cases. Instead the restriction to nodes with complex network topology such as nodes connected to many communication links had network topology dependent disposition and did not achieve good performances in general. As a conclusion, the observation restriction to nodes with large traffic should be adopted if a restriction on the number of observable nodes is imposed for network implementation.

[Related paper]

  1. Shoji Kurakake, ``Control methods for wireless-mesh-network link-scheduling adaptive to traffic changes and application selection adaptive to user activities,'' Ph.D. Thesis, Graduate School of Information Science and Technology, Osaka University, July 2011. [thesis]
  2. Shoji Kurakake, Yasuhiro Ogiri and Hirotaka Nakano, ``An adaptive link scheduling by graph coloring for wireless mesh networks with partial traffic observation capability,'' IEICE Transactions on Information and Systems, vol.J94-D, no.11, pp.1876-1887, Nov. 2011. (in Japanese) [paper]

2.1.5. Effective node repositioning method with limitation of the number of repositioned nodes for IEEE 802.16j relay networks

Effective node repositioning method with limitation of the number of repositioned nodes for IEEE 802.16j relay networks

IEEE 802.16j is now attracting considerable attention as a technology that can provide widearea and high-speed wireless access network service to metropolitan areas or areas in which the construction of wired networks is difficult. In IEEE 802.16j-based multihop relay network, the positions of relay nodes and the network topology are important factors in achieving high performance. In particular, radio wave interference must be considered when determining node positions. However, nodes are generally deployed in limited areas and it is difficult to move them to arbitrary positions. For example, nodes are set up on the roofs of buildings in an urban area and a node can be moved within only the roof area. In wireless sensor networks, various node redeployment methods for such purposes as efficient sensing coverage or network lifetime extension are proposed. In wireless relay networks for high-speed data transmission, however, the transmission efficiency improvement is also important in determining positions of relay nodes.

In this research, we propose a node repositioning method in IEEE 802.16j multihop relay networks to improve network performance. The proposed method works within the constraints such that nodes are repositioned in limited areas. Since the computational cost is high for assessing all possible node positions, the proposed method limits the number of candidate nodes to be repositioned on the basis of distance information. In repositioning multiple nodes, the proposed method also decreases the computational cost by utilizing a heuristic hill-climbing algorithm, since the computational cost becomes much high for trying all combinations of nodes' movement.

We examine the effectiveness of the proposed method through simulation experiments and shows that a performance improvement of up to 28% is realized by repositioning 30% of all nodes. We also find that nearly optimal results are obtained with much lower computational cost than an exhaustive search. Additionally, we exhibit that more performance improvement can be achieved by repositioning many nodes with the proposed method than by repositioning fewer nodes with an exhaustive search.

[Related papers]

  1. Takafumi Shigefuji, ``Effective node repositioning method with limitation of the number of repositioned nodes for IEEE 802.16j relay networks,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.
  2. Takafumi Shigefuji, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``Node repositioning method based on topology information in IEEE 802.16j relay networks,'' in Proceedings of the 8th International Conference on Networking and Services (ICNS 2012), pp.57-62, Mar. 2012. [paper]

2.1.6. Performance evaluation and estimation of IEEE 802.16j relay networks considering radio wave obstruction by obstacles

Performance evaluation and estimation of IEEE 802.16j relay networks considering radio wave obstruction by obstacles

IEEE 802.16j relay networks are expected to provide broadband wireless access environment by using multi-hop wireless communications. In IEEE 802.16j networks, obstacles, which refer in this research to physical objects such as residential, office, and commercial buildings, largely affect the connectivity between relay nodes and radio wave interference among wireless links, and consequently have both advantage and disadvantage on network performance. Therefore, it is important to consider the presence of obstacles and their influence on radio wave propagation for assessing the network performance.

In IEEE 802.16j networks, radio wave interference between wireless links must be taken into account when radio resources are assigned to network links. The protocol model, which defines the transmission and interference ranges as circles, is well-known as one of major radio interference models. Although a lot of related studies on IEEE 802.16j networks use the protocol model, they did not consider the presence of obstacles.

In this research, we investigate the performance of IEEE 802.16j networks considering the effect of obstacles. First, we define an obstacle model which determines location and size of each obstacle in the network. Then, we extend the function of the protocol model for accommodating obstacles, where radio waves propagation is obstructed by obstacles. Next, the performance of IEEE 802.16j networks is evaluated through simulation experiments using the obstacle model and the extended protocol model. Through simulation results, we exhibit that the deployment of additional relay nodes improves the service ratio more effectively than an increase in the radio transmission range of each relay node. We also reveal the influence of the network parameters, such as the number of connected nodes, a transmission range of each node, and distribution density of obstacles, on the performance of IEEE 802.16j networks.

Additionally, we present a method for estimating the performance of IEEE 802.16j networks. When constructing a network, the accurate estimation of the network performance before the network is actually constructed is highly valuable. For that purpose, we use multiple regression analysis based on simulation results and constructs regression equations for network service ratio and power-to-throughput ratio. By evaluating the accuracy of the performed equations based on real-world environment, we confirm that the service ratio can be estimated with a high degree of accuracy when the distribution density of obstacle is small.

[Related papers]

  1. Yuuki Ise, ``Performance evaluation and estimation of IEEE 802.16j relay networks considering radio wave obstruction by obstacles,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.
  2. Yuuki Ise, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``Evaluation of IEEE 802.16j relay network performance considering obstruction of radio wave propagation by obstacles,'' in Proceedings of the 8th International Conference on Networking and Services (ICNS 2012), pp.63-68, Mar. 2012. [paper]

2.2. Research on object counting and tracking

2.2.1. Counting pedestrians in video sequences based on optical flow clustering

Counting pedestrians in video sequences based on optical flow clustering

Pedestrian counting at event sites, buildings, or streets is an essential function for pedestrian traffic control, marketing surveys, surveillance in public places, and so on. Traditionally, pedestrian counting is conducted manually, that entails high cost and it may introduce human errors. Therefore, the demand for automatic counting of pedestrians at various locations has been increased. Numerous approaches to automatically counting pedestrians have been reported in the literature. Among them, pedestrian counting using video-processing technologies has some advantages against sensor-based counting methods, including that the camera position is flexible, the video sequences can be retrieved remotely, and information such as the height and gender of pedestrians can be obtained, in addition to the number of pedestrians. However, the counting accuracy degrades due to many reasons, including occlusion effect where pedestrians are overlapping in video frames.

In this research, we propose a pedestrian counting method using video-processing technologies to improve the counting accuracy under occlusion situations. In the proposed method, we introduce a method of clustering optical flows of moving objects in video frames. The proposed method assumes that optical flows detected from a single pedestrian have similar lengths and angles, and that their starting points exist within a certain range. Therefore, optical flows are clustered based on the lengths, angles, and source locations of optical flows. Although the number of optical flow clusters and the actual number of pedestrians have a strong correlation, the degree of correlation depends on the camera position, the size of pedestrians in video frames, and various factors. Therefore, the proposed method utilizes the pre-learned statistical data to estimate the number of pedestrians.

The counting accuracy of the proposed method is evaluated using several video sequences. we confirm that the proposed method improves the counting accuracy by up to 25% as compared with a non-clustering method. We also report that using a clustering threshold of angles less than 1 is effective for enhancing counting accuracy. Furthermore, we compare the performance of two algorithms which are use feature points and lattice points when optical flows are detected. We confirm that the counting accuracy using feature points is higher than that using lattice points especially when the number of occluded pedestrians increases.

[Related papers]

  1. Shizuka Fujisawa, ``Pedestrian counting in video sequences based on optical flow clustering,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.
  2. Shizuka Fujisawa, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``Pedestrian counting in video sequences using optical flow clustering,'' in Proceedings of the 6th International Conference on Circuits, Systems and Signals (CSS 2012), pp.51-56, Mar. 2012. (Best Paper Award) [paper]
  3. Shizuka Fujisawa, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``Pedestrian counting in video sequences based on optical flow clustering,'' submitted for publication, Mar. 2012.

2.2.2. Counting pedestrians using H.264/AVC video stream

Counting pedestrians using H.264/AVC video stream

We propose a method for estimating the number of pedestrians through statistical analysis of the parameter values generated by the H.264/AVC encoder. A previously reported method estimates the number of pedestrians by using the number of corners extracted from the frame. To perform motion compensation for shape and movement, the H.264/AVC encoder simultaneously extracts values of a number of parameters, such as the number of variable blocks and the residual signal. The proposed method utilizes the extracted parameter values in combination to improve the estimation accuracy. The estimation of the number of pedestrians is compared between three different test videos. In the case that a single parameter is used, the residual signal was the one that gave the most accurate estimate. In this case, roughly the same estimation accuracy was obtained as in the existing method that uses the number of corners. If high-quality video was recorded such that no variable blocks were present in the background, by combining the residual signal with the variable number of blocks, the average absolute error improved by 0.64 persons and the average relative error improved by 7.2%.

[Related paper]

  1. Yasuhiro Ogiri, Shoji Kurakake and Hirotaka Nakano, ``Estimation of number of pedestrians by using H.264/AVC video stream,'' submitted for publication, Feb. 2012. (in Japanese)

2.2.3. Counting pedestrians passing through virtual gate in video sequences based on optical flow clustering

Counting pedestrians passing through virtual gate in video sequences based on optical flow clustering

Pedestrian counting is needed in many situations such as safety control in airports or stations, traffic control in streets and surveillance at event sites. Pedestrian counting is usually conducted manually or automatically using sensors and video sequences captured by cameras. Manual counting requires labor costs and can not be used in crowded scenes due to human error. Pedestrian counting using video sequences has advantage of flexibility of camera placement compared with other sensors, because video sequences can be retrieved remotely. Pedestrian counting systems using video sequences are divided into two kinds of systems. One is used for counting pedestrians for each frame, and the other is used for counting pedestrians passing at specific points or lines. In both pedestrian counting systems, improvement of the counting accuracy under crowded situations is open issue.

In our research group, we have proposed a method for counting pedestrians for each frame. In the method, optical flows which are vectors of moving objects, are first obtained from video sequences. Next, optical flows are clustered according to lengths, angles, and source locations of optical flows. Then, the number of pedestrians are estimated based on the number of clusters. The effectiveness of the method was shown through experimental evaluation using several video sequences. In this research, we propose a method for counting pedestrians passing at a specific line. We call the line virtual gate. The proposed method is based on optical flow clustering like our previous study. Here, there are variations of process to obtain optical flows from video sequences, e.g., coordinate systems used, starting points of optical flows, and the number of frames to obtain optical flows. In this research, we evaluate all variations of process to obtain optical flows, and find the best way to achieve the highest counting accuracy. Through experimental evaluations using a video sequence captured from Cybermedia Center, Osaka University, it was shown that the proposed method achieves higher accuracy when starting points of optical flows are set on the virtual gate.

[Related paper]

  1. Miki Mizushima, ``Counting pedestrians passing through virtual gate in video sequences by extracting optical flows,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)

2.2.4. Counting pedestrians with compound-eye binary sensors

Counting pedestrians with compound-eye binary sensors

In many situations such as survey of the number of visitors to commercial facilities and event sites, there are demands for pedestrian counting. Manual pedestrian counting has some problems such as high-cost and decrease of counting accuracy under crowded situation. Therefore, automated pedestrian counting technologies have attracted much attention. Binary sensors are among the simplest sensors and can detect only the presence or absence of pedestrians within sensing range. Although binary sensors can neither detect the number of pedestrians nor identify pedestrians individually, they have advantages in terms of cost, simplicity, and energy-efficiency. By constructing a wireless sensor network from binary sensors equipped with wireless communication devices, it is possible to construct a pedestrian counting system. However, in the pedestrian counting system with binary sensors, it is a problem that estimation accuracy decreases under crowded situation such as many pedestrians are incoming and outgoing.

In this research, we propose a method for estimating the number of pedestrians under the situation that pedestrians move bidirectionally such as a corridor. In the proposed method, a compound-eye sensor is constructed on a ceiling by closely placing two binary sensors whose sensing regions are partly overlapping. The output data from compound-eye sensor is gathered to a monitoring server through wireless networks, and the monitoring server estimates the number of pedestrians by using the information. In this research, to estimate the number of pedestrians from sensor data, we propose a method based on the Monte Carlo method. Concretely, the proposed method conducts a number of Monte Carlo simulations while the system detects pedestrians. When the system detects no pedestrians, a simulation scenario that has the smallest difference between output histories of compound-eye sensor is selected, and the number of pedestrians under the simulation scenario is used as the estimation result. In this research, we evaluated the fundamental characteristics of the proposed method by changing compound-eye sensor parameters, arrival rate of pedestrians and the number of simulations in the proposed method. In addition, it was shown that the estimation accuracy is improved up to 90% compared with a existing method that estimates the number of pedestrians from difference between output timing of two binary sensors.

[Related paper]

  1. Shuto Fujii, ``Monte Carlo-based bi-directional pedestrian counting with compound-eye binary sensors,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)

2.3. Research on overlay network architectures

2.3.1. Reduction of inter-ISP transit cost caused by overlay routing

Reduction of inter-ISP transit cost caused by overlay routing

Overlay routing is an application-level routing mechanism and existing research has revealed that overlay routing can improve user-perceived performance. On the other hand, overlay routing may harm the ISPs' cost structure because of the policy mismatch between IP routing and overlay routing. In a previous study, we proposed a method to reduce inter-ISP transit cost caused by overlay routing while maintaining the effectiveness of the overlay routing. However, we evaluated the proposed method only in the PlanetLab environment, the node location of which is biased to North America. In addition, the previous evaluation focused only on limiting the degree of increase in the inter-ISP transit cost and did not explicitly consider the performance of the overlay routing. In the present research, we evaluate the proposed method in more general network environments, which are the generalized PlanetLab environment and the Japanese commercial ISPs network environment. We also evaluate the performance of the proposed method in terms of the trade-off relationships between the overlay routing performance and the inter-ISP transit cost over the entire network. In addition, we discuss the differences in the network properties in both environments, which affect the performance of the proposed method.

[Related papers]

  1. Kazuhito Matsuda, Go Hasegawa, Satoshi Kamei and Masayuki Murata, ``Centralized and distributed heuristic algorithms for application-level traffic routing,'' in Proceedings of the 26th International Conference on Information Networking (ICOIN 2012), pp.193-198, Feb. 2012. [paper]
  2. Kazuhito Matsuda, Go Hasegawa, Satoshi Kamei and Masayuki Murata, ``A method to reduce inter-ISP transit cost caused by overlay routing based on end-to-end measurement,'' submitted for publication, Nov. 2011.

2.3.2. Proactive recovery from multiple failures utilizing overlay networking technique

Proactive recovery from multiple failures utilizing overlay networking technique

In this research, we propose a proactive recovery method against multiple network failures for large-scale packet switching networks. The proposed method exploits the overlay networking technique. Specifically, it constructs multiple logical network topologies from the original overlay network topology by assuming various failure patterns. When a failure is detected, our method selects one topology. Consequently, it can immediately recover from the failure by utilizing the selected topology without waiting for routing convergence in the network. When constructing multiple logical topologies, we take into account the correlation among overlay links in terms of the underlay links, and we merge multiple logical topologies into a single topology to decrease the number of topologies while keeping the recovery efficiency. Through the numerical evaluation results of the network reachability and average path length, we show that our method improves network reachability from 69% to 99% while keeping the path length short, when 5% underlay links are simultaneously down.

[Related papers]

  1. Go Hasegawa, Takuro Horie and Masayuki Murata, ``Network recovery method with pre-calculation of routing configurations for large-scale failures,'' Journal of Communications and Networking, vol.1, no.5, pp.60-73, 2011.
  2. Go Hasegawa, Takuro Horie and Masayuki Murata, ``Proactive recovery from multiple failures utilizing overlay networking technique,'' Telecommunication Systems, 2012. [paper]

2.3.3. Scalable and density-aware measurement strategies for overlay networks

Scalable and density-aware measurement strategies for overlay networks

In network performance measurement in overlay networks, measurement conflict occurs when multiple overlay paths share the underlay network route when they are measured at the same time, which increase the network load and the measurement accuracy degrades. We have proposed the spatial composition method of measurement results, where we estimate the network performance of the longer overlay path from the measurement results of partial paths, when the longer path includes other overlay nodes. In this report, the author evaluates the accuracy of the composition method in terms of packet loss ratio, by utilizing the measurement results in PlanetLab environment. From the evaluation results, we confim that the estimation error is enough small.

[Related papers]

  1. Dinh Tien Hoang, Go Hasegawa and Masayuki Murata, ``A distributed measurement method for reducing measurement conflict frequency in overlay networks,'' in Proceedings of the 2011 International Communications Quality and Reliability Workshop (IEEE CQR 2011), pp.1-6, May 2011. [paper]
  2. Yusuke Iijima, ``Spatial composition method for packet loss ratio measurement on overlay networks,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)

2.3.4. The implication of overlay routing on ISPs' connecting strategies

The implication of overlay routing on ISPs' connecting strategies

The Internet is composed of many distinct networks, operated by independent Internet Service Providers (ISPs). There are primarily two kinds of relationships between ISPs: transit and peering. ISPs' traffic and economic relationships are mainly decided by ISPs' routing policy. However, in today's Internet, overlay routing, which changes traffic routing at the application layer to better satisfy the applications' demands, is rapidly increasing, and brings challenge to the ISPs' settlement interconnection researches. The goal of this research is to study the economic implications of overlay routing on ISPs' peering incentive, costs and strategy choice. For this purpose, we introduce an ISP interconnection business model based on a simple ISPs' network. We then study the overlay traffic patterns in the network in various conditions. Combining the business model with traffic patterns, we study the ISPs' economic issues such as incentive to upgrade peering link and cost reduction conditions with various overlay traffic patterns and settlement methods. AT last, we analyze the bilateral Nash equilibrium (BNE) strategy of ISPs in the network. We also give some numerical examples to explain our results.

[Related papers]

  1. Xun Shao, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``The implication of overlay routing on ISPs' connecting strategies,'' in Proceedings of the 23rd International Teletraffic Congress (ITC 2011), pp.286-293, Sep. 2011. [paper] [slide]

2.4. Research on architecture of transport layer for next-generation high-speed network

2.4.1. Analysis of power consumption in TCP data transmission over a wireless LAN environment

Analysis of power consumption in TCP data transmission over a wireless LAN environment

Reduction of power consumption is one of an important issue in wireless communications because most mobile devices are battery-driven. For effective power saving, it is important to understand the behaviors of the transport layer protocols used by upper-layer applications, since packet transmission and reception timing are mainly determined by those behaviors. In this report, we propose a power consumption model based on a treatment of detailed TCP behaviors within a WLAN environment. Comparing the model with and without ideal sleeping, we analyze the power consumption of a single wireless client as it sends data to a wired host by TCP. From the numerical results of our analysis, we show the lower-bound for power consumption in upstream TCP data transfer with ideal sleeping. We also discuss the trade-off between power saving and network performance in TCP data transfer.

[Related papers]

  1. Masafumi Hashimoto, Go Hasegawa and Masayuki Murata, ``Modeling and analysis of power consumption in TCP data transmission over a wireless LAN environment,'' in Proceedings of the 4th International Workshop on Green Communications (GreenComm 2011), pp.1-6, June 2011. [paper]
  2. Masafumi Hashimoto, Go Hasegawa and Masayuki Murata, ``Energy efficiency analysis of TCP with burst transmission over a wireless LAN,'' in Proceedings of the 11th IEEE International Symposium on Communications and Information Technologies (IEEE ISCIT 2011), pp.292-297, Oct. 2011. [paper]

2.4.2. Influence of power saving technique in IEEE 802.11 multi-hop networks on network performance

Influence of power saving technique in IEEE 802.11 multi-hop networks on network performance

The issue of power consumption in wireless networks is becoming increasingly important due to the rapid development of various wireless devices such as sensors, smartphones, and tablet PCs. The IEEE 802.11 wireless LAN standard defines multiple data transmission protocols, each of which has various characteristics such as power consumption, data rate, modulation method, and transmission distance. Therefore, it is important to choose the optimal data rate in terms of power consumption as well as throughput, especially when considering data transmission over multi-hop networks.

In this research, we present a mathematical analysis of power consumption in data transmission over IEEE 802.11-based wireless multi-hop networks to investigate the effect of data rate selection on power consumption. The analysis results show that there are some situations where a low data rate should intentionally be selected in order to minimize power consumption. Our analysis indicates that power consumption can be decreased by up to 13% when the symbol error rate is comparatively small.

[Related paper]

  1. Wataru Toorisaka, Go Hasegawa and Masayuki Murata, ``Power consumption analysis of data transmission in IEEE 802.11 multi-hop networks,'' in Proceedings of the 8th International Conference on Networking and Services (ICNS 2012), pp.75-80, Mar. 2012. [paper]

2.4.3. Improvement of end-to-end bandwidth measurement method for power-saving routers

Improvement of end-to-end bandwidth measurement method for power-saving routers

The increase in energy consumption associated with network traffic explosion is becoming a major problem. To realize energy-efficient networking, a number of researchers have focused on technologies that dynamically adjust the processing performance and the link speed of routers and switches according to the network traffic load. However, when such power-saving routers and switches are present in the network, the existing network protocols and control methods may not work well since they do not assume the existence of such routers and switches. We focuses on the measurement method of an end-to-end available bandwidth, which is an important performance metric for various network controls.

In this research, we introduce a parameter tuning method for Pathload, which is a popular tool for measuring the end-to-end available bandwidth, to maintain the measurement accuracy in the network with power-saving routers, while not affecting the behavior of power-saving routers. For that purpose, we first investigate the interactions between the bandwidth measurement behavior of Pathload and the energy efficiency of power-saving routers by conducting extensive simulation experiments. In the simulation experiments, we examine the effects of the power-saving router on the measurement accuracy and the effects of the Pathload measurement on the behavior of the power-saving router. Based on simulation results, we exhibit that both the measurement accuracy of Pathload and the energy efficiency of power-saving routers degrade, particularly when the power saving functions are triggered in short cycles.

As the second contribution of this research, we propose a parameter tuning method for Pathload that maintain measurement accuracy without affecting the behavior of power-saving routers. We accomplish this by giving simple mathematical analysis of the traffic load of Pathload on the power-saving router. We confirm the accuracy of our analysis by comparing the analysis results with corresponding simulation results. We also show that Pathload can measure an available bandwidth without affecting the behavior of power-saving routers by tuning the number of probing packets on the basis of the analysis results.

[Related paper]

  1. Daisuke Kobayashi, ``Parameter tuning of end-to-end bandwidth measurement method with power-saving routers,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.
  2. Daisuke Kobayashi, Go Hasegawa and Masayuki Murata, ``Evaluation and improvement of end-to-end bandwidth measurement method for power-saving routers,'' in Proceedings of the IEEE 2011 International Communications Quality and Reliability Workshop (IEEE CQR 2011), May 2012. (to be presented)

2.4.4. Measuring available bandwidth of multiple parts on end-to-end network path

Measuring available bandwidth of multiple parts on end-to-end network path

The available bandwidth on an end-to-end network path is an important performance metric for detecting network congestion, adapting transmission rate, configuring path and topology on overlay networks, and so on. The existing available bandwidth measurement techniques aimed only at knowing available bandwidth of the bottleneck part on the path and they do not specify where is the bottleneck. Also, they can not measure available bandwidth of multiple parts on the path.

In this research, we propose a simultaneous measurement method of available bandwidth of multiple parts on an end-to-end network path. The proposed method estimates the available bandwidth based on changes in packet arrival intervals under the situation where intermediate routers can make timestamps of incoming packets. We present extensive simulation results of the proposed method and confirm that it can accurately measure available bandwidth of each part on the path even when the available bandwidth of the sender side network is smaller than that of the receiver side network.

[Related paper]

  1. Kazumasa Koitani, Go Hasegawa and Masayuki Murata, ``Measuring available bandwidth of multiple parts on end-to-end network path,'' in Proceedings of the IEEE 2012 International Communications Quality and Reliability Workshop (IEEE CQR 2012), May 2012. (to be presented)

2.4.5. Design and evaluation of a data transfer mechanism for video streaming over TCP

Design and evaluation of a data transfer mechanism for video streaming over TCP

In recent years, video streaming services using TCP as a transport layer protocol represented by YouTube have been popular, with rapid increase of the network bandwidth. Congestion control of TCP increases the congestion window until a packet loss is detected. Therefore, the video streaming over TCP sends packets to a network at a higher speed than the video playback speed that takes the bandwidth of other competing traffic unnecessarily.

In this research, we first investigate the data transfer mechanism of existing video streaming services using TCP and show that the existing video streaming services perform data transfer at the extremely higher rate than the video playback speed. Then, we propose the data transfer mechanism for video streaming over TCP that does not take the network bandwidth excessively by controlling the data transfer at an application layer.

The proposed mechanism acquires the TCP state variables from the source host TCP, and estimates network congestion status. Moreover, the amount of buffered video data is acquired from a destination host. By adjusting the amount of data passed to a source host TCP in one round-trip time based on the network congestion status and the amount of buffered video data, the proposed mechanism controls the data transfer speed. Simulation results show that the proposed mechanism decreases the frequency of buffer underflow at the destination host, and does not take the bandwidth of background traffic.

[Related paper]

  1. Hiroyuki Hisamatsu, Go Hasegawa and Masayuki Murata, ``Non bandwidth-intrusive video streaming over TCP,'' in Proceedings of the 8th International Conference on Information Technology: New Generations (ITNG 2011), pp.78-83, Apr. 2011. [paper]
  2. Hiroyuki Hisamatsu, Go Hasegawa and Masayuki Murata, ``Network friendly transmission control for progressive download over TCP,'' Journal of Communications, vol.6, no.3, pp.213-221, Mar. 2012. [paper]

2.4.6. Improving interactive user experience on thin clients

Improving interactive user experience on thin clients

Return network traffic (from servers to clients) in thin-client systems is modeled as a mixture of interactive data flows corresponding to keystrokes and bulk data flows related to screen updates. Users are very sensitive to delay and jitter of the former flows. Thus our goal is to minimize the latency of interactive data transfer without increasing latency of bulk data transfer. Through simulation experiments, we determine that the main factors causing end-to-end delay in the interactive data transfer are queuing delay in the router and buffering delay in the server. When we apply two techniques: priority queuing of interactive data flows at the router and using TCP SACK option, the average end-to-end delay can be reduced. However, several servers could take more than a second to send large bulk data flows; this delays the transmission of following interactive data flows. We then develop TCP optimization mechanisms: modifying recalculation of the retransmission timeout value and temporarily turning off the TCP SACK control, and demonstrate that they can overcome the negative effects of the existing techniques.

[Related paper]

  1. Yukio Ogawa, Go Hasegawa and Masayuki Murata, ``A transport-layer approach for improving thin-client performance in a WAN environment,'' International Journal of Internet Protocol Technology (IJIPT), vol.6, no.3, pp.172-183, Nov. 2011. [paper]

2.4.7. Performance evaluation of distributed computing environment considering power consumption

Performance evaluation of distributed computing environment considering power consumption

When information systems are consolidated in a large data center, a huge amount of energy is consumed for transferring massive amounts of data to the data center through a wide area network. Our approach is to reduce the network power consumption by distributing servers capable of reducing or localizing traffic across the network. In this research, we develop the models of applications, their traffic and a network system for transferring the traffic. We also introduce the models of router's and server's energy consumption, and formulate the power consumption caused by the traffic. Then, we evaluate the power consumption reduction as a result of traffic reduction by the distributed servers. Numerical evaluation results show that the whole power consumption reduces by about 20 percent when the more servers are added to the network; it increases due to the server power consumption when a lot of servers are added to the network.

[Related papers]

  1. Yukio Ogawa, ``Solution approaches for wide-area distributed systems toward integration of enterprise networks and computing resources,'' Ph.D. Thesis, Graduate School of Information Science and Technology, Osaka University, Jan. 2012.
  2. Yukio Ogawa, Go Hasegawa and Masayuki Murata, ``Power consumption evaluation of distributed computing network considering traffic locality,'' submitted for publication, Nov. 2011.
  3. Yukio Ogawa, Go Hasegawa and Masayuki Murata, ``Effect of traffic locality on power consumption of distributed computing network with a data center,'' submitted for publication, Nov. 2011.

3. Publications

3.1. Book chapters

  1. Go Hasegawa and Masayuki Murata, ``TCP symbiosis: Bio-inspired congestion control mechanism for TCP,'' in Biologically Inspired Networking and Sensing: Algorithms and Architectures, IGI Global, pp.104-131, Aug. 2011. [paper]

3.2. Journal papers

  1. Go Hasegawa and Masayuki Murata, ``Research trends on TCP congestion control mechanisms,'' IEICE Transactions on Communications, vol.J94-B, no.5, pp.663-672, May 2011. (Invited) (in Japanese) [paper]
  2. Masakazu Murata, Yoshiaki Taniguchi, Go Hasegawa and Hirotaka Nakano, ``SHOT: Scenario-type hypothesis object tracking with indoor sensor networks,'' IEICE Transactions on Information and Systems, vol.E94-D, no.5, pp.1035-1044, May 2011. [paper]
  3. Yoshiaki Taniguchi, Tomoya Kitani and Kenji Leibnitz, ``A uniform airdrop deployment method for large-scale wireless sensor networks,'' International Journal of Sensor Networks (IJSNet), vol.9, no.3-4, pp.182-191, May 2011. [paper]
  4. Takayuki Nishio, Ryoichi Shinkuma, Tatsuro Takahashi and Go Hasegawa, ``TCP-based window-size delegation method for TXOP exchange in wireless local area networks,'' EURASIP Journal on Wireless Communications and Networking, vol.2011, article ID 2011:58, 12 pages, Aug. 2011. [paper]
  5. Go Hasegawa, Toshiaki Hidekuma, Masahiro Sasabe and Hirotaka Nakano, ``Power control methods for improving spatial reuse in TDMA-based wireless mesh networks,'' ISRN Communications and Networking, vol.2011, article ID 970943, 12 pages, 2011. [paper]
  6. Shoji Kurakake, Yasuhiro Ogiri and Hirotaka Nakano, ``An adaptive link scheduling by graph coloring for wireless mesh networks with partial traffic observation capability,'' IEICE Transactions on Information and Systems, vol.J94-D, no.11, pp.1876-1887, Nov. 2011. (in Japanese) [paper]
  7. Gyeongyeon Kang, Yoshiaki Taniguchi, Go Hasegawa and Hirotaka Nakano, ``Parameter tuning of the protocol interference model using SINR for time slot assignment in wireless mesh networks,'' IEICE Transactions on Information and Systems, vol.E94-D, no.11, pp.2191-2200, Nov. 2011. [paper]
  8. Yukio Ogawa, Go Hasegawa and Masayuki Murata, ``A transport-layer approach for improving thin-client performance in a WAN environment,'' International Journal of Internet Protocol Technology (IJIPT), vol.6, no.3, pp.172-183, Nov. 2011. [paper]
  9. Go Hasegawa, Takuro Horie and Masayuki Murata, ``Network recovery method with pre-calculation of routing configurations for large-scale failures,'' Journal of Communications and Networking, vol.1, no.5, pp.60-73, 2011.
  10. Yoshiaki Taniguchi, Akimitsu Kanzaki, Naoki Wakamiya and Takahiro Hara, ``An energy-efficient data gathering mechanism using traveling wave and spatial interpolation for wireless sensor networks,'' Journal of Information Processing (JIP), vol.20, no.1, pp.167-176, Jan. 2012. [paper]
  11. Hiroyuki Hisamatsu, Go Hasegawa and Masayuki Murata, ``Network friendly transmission control for progressive download over TCP,'' Journal of Communications, vol.6, no.3, pp.213-221, Mar. 2012. [paper]
  12. Go Hasegawa, Rintaro Ishii, Yoshiaki Taniguchi and Hirotaka Nakano, ``Time slot assignment algorithms for reducing transmission latency in IEEE 802.16j multi-hop relay networks,'' Journal of Communications and Networking, vol.2, no.3, Mar. 2012.

3.3. International conference papers

  1. Hiroyuki Hisamatsu, Go Hasegawa and Masayuki Murata, ``Non bandwidth-intrusive video streaming over TCP,'' in Proceedings of the 8th International Conference on Information Technology: New Generations (ITNG 2011), pp.78-83, Apr. 2011. [paper]
  2. Dinh Tien Hoang, Go Hasegawa and Masayuki Murata, ``A distributed measurement method for reducing measurement conflict frequency in overlay networks,'' in Proceedings of the 2011 International Communications Quality and Reliability Workshop (IEEE CQR 2011), pp.1-6, May 2011. [paper]
  3. Takayuki Nishio, Ryoichi Shinkuma, Tatsuro Takahashi and Go Hasegawa, ``TCP window-size delegation for TXOP exchange in wireless access networks,'' in Proceedings of IEEE International Conference on Communications (IEEE ICC 2011), pp.1-5, June 2011. [paper]
  4. Masafumi Hashimoto, Go Hasegawa and Masayuki Murata, ``Modeling and analysis of power consumption in TCP data transmission over a wireless LAN environment,'' in Proceedings of the 4th International Workshop on Green Communications (GreenComm 2011), pp.1-6, June 2011. [paper]
  5. Xun Shao, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``The implication of overlay routing on ISPs' connecting strategies,'' in Proceedings of the 23rd International Teletraffic Congress (ITC 2011), pp.286-293, Sep. 2011. [paper] [slide]
  6. Masafumi Hashimoto, Go Hasegawa and Masayuki Murata, ``Energy efficiency analysis of TCP with burst transmission over a wireless LAN,'' in Proceedings of the 11th IEEE International Symposium on Communications and Information Technologies (IEEE ISCIT 2011), pp.292-297, Oct. 2011. [paper]
  7. Kazuhito Matsuda, Go Hasegawa, Satoshi Kamei and Masayuki Murata, ``Centralized and distributed heuristic algorithms for application-level traffic routing,'' in Proceedings of the 26th International Conference on Information Networking (ICOIN 2012), pp.193-198, Feb. 2012. [paper]
  8. Shizuka Fujisawa, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``Pedestrian counting in video sequences using optical flow clustering,'' in Proceedings of the 6th International Conference on Circuits, Systems and Signals (CSS 2012), pp.51-56, Mar. 2012. (Best Paper Award) [paper]
  9. Takafumi Shigefuji, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``Node repositioning method based on topology information in IEEE 802.16j relay networks,'' in Proceedings of the 8th International Conference on Networking and Services (ICNS 2012), pp.57-62, Mar. 2012. [paper]
  10. Yuuki Ise, Go Hasegawa, Yoshiaki Taniguchi and Hirotaka Nakano, ``Evaluation of IEEE 802.16j relay network performance considering obstruction of radio wave propagation by obstacles,'' in Proceedings of the 8th International Conference on Networking and Services (ICNS 2012), pp.63-68, Mar. 2012. [paper]
  11. Wataru Toorisaka, Go Hasegawa and Masayuki Murata, ``Power consumption analysis of data transmission in IEEE 802.11 multi-hop networks,'' in Proceedings of the 8th International Conference on Networking and Services (ICNS 2012), pp.75-80, Mar. 2012. [paper]

3.4. Theses

3.4.1. Ph.D. Thesis

  1. Shoji Kurakake, ``Control methods for wireless-mesh-network link-scheduling adaptive to traffic changes and application selection adaptive to user activities,'' Ph.D. Thesis, Graduate School of Information Science and Technology, Osaka University, July 2011. [thesis]

3.4.2. Masters' Theses

  1. Yuuki Ise, ``Performance evaluation and estimation of IEEE 802.16j relay networks considering radio wave obstruction by obstacles,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.
  2. Daisuke Kobayashi, ``Parameter tuning of end-to-end bandwidth measurement method with power-saving routers,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.
  3. Takafumi Shigefuji, ``Effective node repositioning method with limitation of the number of repositioned nodes for IEEE 802.16j relay networks,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.
  4. Shizuka Fujisawa, ``Pedestrian counting in video sequences based on optical flow clustering,'' Master's Thesis, Graduate School of Information Science and Technology, Osaka University, Feb. 2012.

3.4.3. Bachelors' Theses

  1. Yusuke Iijima, ``Spatial composition method for packet loss ratio measurement on overlay networks,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)
  2. Shoichi Takagi, ``Time slot assignment method based on estimation of radio wave interference in IEEE 802.16j relay networks,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)
  3. Kohei Higo, ``Time slot assignment mechanism considering adaptive modulation and coding in IEEE 802.16j relay networks,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)
  4. Shuto Fujii, ``Monte Carlo-based bi-directional pedestrian counting with compound-eye binary sensors,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)
  5. Miki Mizushima, ``Counting pedestrians passing through virtual gate in video sequences by extracting optical flows,'' Bachelor's Thesis, School of Engineering Science, Osaka University, Feb. 2012. (in Japanese)