Gergely Vakulya

Fast Adaptive Acoustic Localization for Sensor Networks

Sensor networks have proven their usefulness in various acoustic localization applications. Recently, a consistency-function-based algorithm has been proposed, which can provide accurate solutions even if a large number of independent outliers are present in a measurement set. In certain practical cases, e.g., in non-line-of-sight reverberant areas, however, sensors may have cooperative and consistent errors, resulting in bad estimates. In this paper, an adaptive consistency-function-based solution is proposed, which can compensate for cooperative and systematic measurement errors and thus provides accurate results even if the original consistency-function-based algorithm fails. Stochastic initialization is also proposed, which is able to accelerate execution of the algorithm by several orders of magnitude while the global optimum is still provided with arbitrarily high probability.

Design of a sensor network based security system

In this paper a wireless distributed building safety and security framework will be introduced. The main design goals of the system are completely distributed operation, flexibility, robustness, fast response time, and energy efficiency. The framework uses low duty cycle TDMA-based communication scheme, where the scheduling is the result of optimization, based on network discovery measurements. The hardware and software architecture of a prototype system is introduced.

Energy-efficient and reliable round-robin TDMA for wireless sensor networks

In this paper a fault-tolerant energy-efficient Time Division Multiple Access (TDMA) based routing scheme is proposed for wireless sensor networks. The proposed system provides guaranteed delivery times and tolerates node and link failures. The software architecture and the performance properties of the system are described.

Extended round-robin TDMA scheduling scheme for wireless sensor networks

In this paper a TDMA based routing scheme is proposed for such wireless sensor networks, where nodes form multiple connected cycles. The proposed method provides fault tolerance, guaranteed delivery times, and low power consumption. The paper discusses the routing scheme and its implementation details, and provides several performance tests.

Lookup: Robust and Accurate Indoor Localization Using Visible Light Communication

A novel indoor localization system is presented, where LED beacons are utilized to determine the position of the target sensor, including a camera, an inclinometer, and a magnetometer. The beacons, which can be a part of the existing lighting infrastructure, transmit their identifiers for long distances using visible light communication techniques. The sensor is able to sense and detect the high-frequency (flicker free) code by properly undersampling the transmitted signal. The localization is performed using novel geometric and consensus-based techniques, which tolerate well measurement inaccuracies and sporadic outliers. The performance of the system is analyzed using simulations and real measurements. According to large-scale tests in realistic environments, the accuracy of the proposed system is in the low decimeter range.

Adaptive acoustic localization algorithm for sensor networks

Sensor networks have proved their usefulness in various acoustic localization applications. Recently a consistency function-based algorithm was proposed, which can provide accurate solution even if a large number of independent outliers are present in the measurement set. In certain practical cases, however, sensors may have cooperative and consistent errors, resulting in bad estimations. In this paper an adaptive consistency function-based solution is proposed, which can compensate for co-operative and systematic measurement errors and thus provides accurate results even if the original consistency function-based algorithm fails.

Forest intrusion detection system with sensor network

An intrusion detection system was developed to detect illegal entries to guarded forest areas and thus help the prevention of timber theft and other damages, caused by illegal activities. The system contains accelerometers, magnetometers and geophones to provide high detection reliability and fast reaction time. The performance of the proposed system is illustrated by real measurements.

Low power accelerometer based intrusion and tamper detector

In this paper a novel sensor is proposed, which can be applied both to intrusion detection or tamper detection purposes. The device senses vibrations caused by the attacker, using MEMS accelerometers, to provide alarm signals. The sensor uses low duty cycling operation to conserve its energy, thus it is suitable for wireless applications. The performance of the sensor is illustrated with real measurements.

Optimal Multi-TDMA Scheduling in Ring Topology Networks

A scheduling algorithm will be proposed for wireless ring topology networks, utilizing time division multiple access (TDMA) with possible simultaneous operation of nodes. The proposed algorithm finds the optimal schedule to minimize the turnaround time for messages in the network. The properties of the algorithm are mathematically analyzed and proven, and practical test results are also provided.

Low-power communication protocol for low duty cycle data acquisition applications

In this paper a low-power communication protocol is proposed for sensor networks, where sensors mostly send measurements and rarely receive commands. The proposed protocol optimizes energy utilization on sensor nodes by using a modified MAC-layer, with acknowledgment frames piggy-backing meta-information towards sensor nodes. The proposed method is evaluated using real measurements using a modified protocol stack of IEEE 802.15.4.