1,720,970 research outputs found
Effects of Colored Noise in Linear Adaptive Filters Applied to GNSS Multipath Detection
No full textMultipath often represents the dominant error source in GNSS (Global Navigation Satellite System). As a consequence, several techniques have been proposed to mitigate its undesiredeffects, including Linear Adaptive Filters (LAFs). In this paper a Least Squares (LS) filter is applied to a post-correlation GNSS signal. The case study considers a Galileo signal. Since after the correlation the noise results not to be white, which violate one of the basic assumptions of LAF theory, the focus of this paper is on the analysis of the colored noise effects on the LAF performance. An analytical study is done, proved by simulation results
Multipath distance detector algorithm (MPDD): Enhancement and application to Galileo signals
No full text
Effects of Noise Correlation on LEAST SQUARES Filtering in Multipath Detection for GNSS
Full text linkPublication in the conference proceedings of EUSIPCO, Marrakech, Morocco, 201
Key Management in Wireless Sensor Networks, IP-Based Sensor Networks, Content Centric Networks
Full text linkCryptographic keys and their management in network communication is considered the main building block of security over which other security primitives are based. These cryptographic keys ensure the privacy, authentication, integrity and non-repudiation of messages. However, the use of these cryptographic keys and their management in dealing with the resource constrained devices (i.e. Sensor nodes) is a challenging task. A number of key management schemes have been introduced by researchers all over the world for such resource constrained networks. For example, light weight PKI and elliptic curve cryptography schemes are computationally expensive for these resource constrained devices. So far the symmetric key approach is considered best for these constrained networks and different variants of it been developed for these networks (i.e. probabilistic key distribution approach). The probabilistic key distribution approach consumes less memory than the standard symmetric key approach but it suffers from the connectivity issues (i.e. the connectivity depends on the common shared keys between the nodes). Most of those schemes were proposed by considering static sensor networks (e.g. Industrial process monitoring, Environmental monitoring, movement detection in military applications, forests etc.). However, the use of these existing key management schemes for mobile wireless sensor networks applications introduces more challenges in terms of network connectivity, energy consumption, memory cost, communication overhead and protection of key materials against some well known attacks. Keeping these challenges in mind, previous research has proposed some key management schemes considering the mobility scenarios in ad hoc networks and wireless sensor networks (e.g. vehicular networks, health monitoring systems).However these schemes consume more resource because of a much higher communication packet exchange during the handover phase for the authentication of joining and leaving nodes than the static networks where there is no extra communication for the handover and authentication. The motivation of this research work is to investigate and propose new algorithms not only to improve the efficiency of these existing authentication and key management schemes in terms of connectivity, memory and security by considering the mobility scenario in wireless sensor networks, but also to develop new algorithms that suit these constrained networks than the existing schemes. First, we choose the existing key pool approach for authentication and key management and improve its network connectivity and resilience against some well known attacks (e.g. node capturing attacks) while reduce the memory cost by storing those key pools in each sensor node. In the proposed solution, we have divided the main key pool into two virtual mutually exclusive key pools. This division and constructing a key from two chosen keys, one from each key pool, helps to reduce the memory cost of each node by assigning fewer keys for the same level of network connectivity as the existing key pool frameworks. Although, the proposed key pool approach increases the network resilience against node compromission attacks because of the smaller number of keys assigned to each node, however it does not completely nullify the effect of the attacks. Hence we proposed an online mutual authentication and key establishment and management scheme for sensor networks that provides almost 100\% network connectivity and also nullifies the effect of node compromission attacks. In the proposed online key generation approach, the secret key is dependent on both communicating parties. Once the two communicating parties authenticate each other, they would successfully establish a secret communication key, otherwise they stop communication and inform the network manager about the intruder detection and activity. The last part of the thesis considers the integration of two different technologies (i.e. wireless sensor networks and IP networks). This is a very interesting and demanding research area because of its numerous applications, such as smart energy, smart city etc.. However the security requirements of these two kind of networks (resource constrained and resourceful) make key management a challenging task. Hence we use an online key generation approach using elliptic curve cryptography which gives the same security level as the standard PKI approach used in IP networks with smaller key length and is suited for the sensor network packet size limitations. It also uses a less computationally expensive approach than PKI and hence makes ECC suitable to be adopted in wireless sensor networks. In the key management scheme for IP based sensor networks, we generate the public private key pair based on ECC for each individual sensor node. However the public key is not only dependent on the node's parameter but also the parameters of the network to which it belongs. This increases the security of the proposed solution and avoids intruders pretending to be authentic members of the network(s) by spreading their own public keys. In the last part of the thesis we consider Content Centric Networking (CCN) which is a new routing architecture for the internet of the future. Building on the observation that today's communications are more oriented towards content retrieval (web, P2P, etc.) than point-to-point communications (VoIP, IM, etc.), CCN proposes a radical revision of the Internet architecture switching from named hosts (TCP/IP protocols) to named data to best match its current usage. In a nutshell, content is addressable, routable, self-sufficient and authenticated, while locations no longer matter. Data is seen and identified directly by a routable name instead of a location (the address of the server). Consequently, data is directly requested at the network level not from its holder, hence there is no need for the DNS). To improve content diffusion, CCN relies on data distribution and duplication, because storage is cheaper than bandwidth: every content - particularly popular one - can be replicated and stored on any CCN node, even untrustworthy. People looking for particular content can securely retrieve it in a P2P-way from the best locations available. So far, there has been little investigation of the security of CCNs and there is no specific key management scheme for that. We propose an authentication and key establishment scheme for CCNs in which the contents are authenticated by the content generating node, using pre-distributed shares of encryption keys. The content requesting node can get those shares from any node in the network, even from malicious and intruder ones, in accordance with a key concept of CCNs. In our work we also provide means to protect the distributed shares from modification by these malicious/intruder nodes. The proposed scheme is again an online key generation approach but including a relation between the content and its encryption key. This dependency prevents the attackers from modifying the packet or the key share
Total Electron Content measurements with uncertainty estimate
No full textHigh-accuracy Total Electron Content (TEC) measurements are important for several applications, including GNSS time transfer with accuracies better than 10 ns. In addition, it is important to provide uncertainty estimates to algorithms that use the TEC measurements. Dual-frequency TEC errors are dominated by satellite inter-frequency biases, receiver inter-frequency biases, receiver noise and multipath, higher-order ionospheric errors, and receiver inter-code biases for certain receivers. To evaluate different algorithms and error mitigation methods, data were collected from four different GPS receivers in two locations at the Arecibo Observatory in Puerto Rico. Error mitigation methods include receiver hardware bias calibration, comparison with TEC maps provided by the Jet Propulsion Laboratory (JPL), receiver comparisons, TEC calibration during low TEC count as determined by the Arecibo radar, and multipath mitigation using the Code Noise Multi-Path (CNMP) algorithm. The CNMP algorithm enables the definition of rigorous multipath error bounds on the TEC bias estimatio
Multipath Mitigation Using Linear Adaptive Filtering Techniques
No full textLinear adaptive filters result to be suitable to treat the Global Navigation Satellite System (GNSS) multipath channel, given that the signal affected by multipath results as a weighted sum of delayed replicas of a same signal. Since the multipath is becoming more and more one of the limiting factor of the GNSS performance, the problem of its detection and mitigation is one of the most discussed topics in literature. In thi paper a linear adaptive filter, in particular, a Least Squares (LS) filter is applied in post-correlation to a Global Positioning System (GPS) signal. The aim is to find an estimate of the distortion introduced by the multipath on the correlation function, in order to correct the code-delay estimate. Moreover, an analysis is done in order to study the effects of the number of samples in the correlation function. In fact the more correlators are considered, the higher accurate the estimate can be, but the higher the computational effort is required. The aim of this analysis is therefore to find a trade-off in terms of performance and complexity. The proposed estimation method is presented and simulation results are shown, where the GPS data are simulated with a signal simulato
Laboratory and Flight Test Analysis of Rubidium Frequency Reference Performance
No full textThe performance of Rubidium frequency standards is well-documented for stationary operation. In dynamic environments that include changes in orientation, acceleration, temperature, pressure, and magnetic field, the performance of frequency references for position, velocity, and timing applications is not well characterized in the literature. This paper includes results from laboratory and flight test experiments. Laboratory tests were conducted to characterize Rubidium oscillator performance under varying gravity and magnetic field conditions. Three Rubidium oscillators were flight tested. Each of the oscillators was connected to a GPS receiver for error characterization. The frequency stability performance is profiled against the flight dynamics based on a navigation-grade inertial navigation system. The goal of the paper is to increase the awareness of Rb oscillator error sources that affect in-flight frequency stability, with an emphasis on gravity sensitivity and magnetic field susceptibilit
- …
