http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Maximal Connectivity Local Routing for Self-Recovering Wireless Sensor Network
Odilbek Urmonov,Saurabh Kumar,HyungWon Kim(김형원) 대한전자공학회 2016 대한전자공학회 학술대회 Vol.2016 No.6
Wireless sensor networks of mesh structure can experience serious network outage or performance degradation due to node failures. This paper proposes a fail recovery method for multi-hop wireless sensor networks based on TDMA. It prepares for a potential recovery by selecting a backup parent node for each sub-tree. When a node fails, it locally reroutes the lost connections for the child nodes of the faulty node via the backup node. It can, therefore, locally recover the failed connections without rerouting the entire network or re-allocating TDM time slots – a key advantage of the proposed method. We have implemented the proposed algorithm in a simulator. Simulation results show that the proposed recovery method provides substantial improvement in the packet delivery ratio and network lifetime compared to a conventional method.
Odilbek Ikromjonovich Urmonov 대한전자공학회 2008 ICEIC:International Conference on Electronics, Inf Vol.1 No.1
This paper describes a hybrid network implementation that uses both ad hoc connectivity and access points.The network also allows mobile hosts that are multiple hops from an access point to use centralized services, likeHFC, which are not available in pure ad hoc networks.Many scenarios may benefit from this extension of services, such as mobile users near university buildings or at an airport. For efficiency, the “radius” of an access point is limited to K hops. This means all routes have at most K consecutive wireless hops before reaching the destination or an access point. We believe this limitationmay lead to more efficient routing by trading off someconnectivity. The protocol uses proactive routing atthe access points and ondemand routing at the mobilehosts. We present an implementation done as;klproof-of-concept and a basis for future research.
Self-Supervised Training Method of Vehicle Detection CNN Models
Odilbek Urmonov,HyungWon Kim 한국차세대컴퓨팅학회 2022 한국차세대컴퓨팅학회 학술대회 Vol.2022 No.10
Autonomous driving relies on an accurate perception system that provides knowledge about surroundings and ensures safe driving performance. Usually, the perception system takes input information from onboard sensors (camera, LIDAR, RADAR, etc.) and then uses it to perform object detection tasks to accurately determine objects such as pedestrians, vehicles, traffic signs, and road barriers located around the ego vehicle. In order to have a safe trip and maneuver on the road, a vehicle detection algorithm should constantly improve the accuracy of vehicle detection. Since most of the conventional deep learning methods for vehicle or object detection rely on offline training with human-labeled large datasets, the conventional training methods have serious limitations in developing a breakthrough technique for gradual improvement in the detection accuracy of deep learning models. Thus, we propose a self-supervised training (SST) scheme that can gradually enhance detection accuracy with pseudo labeling.