Copyright © All rights reserved micansinfotech.com
For target tracking applications, wireless sensor nodes provide accurate information since they can be deployed and operated near the phenomenon. These sensing devices have the opportunity of collaboration among themselves to improve the target localization and tracking accuracies. An energy-efficient collaborative target tracking paradigm is developed for wireless sensor networks (WSNs). In addition, a novel approach to energy savings in WSNs is devised in the information-controlled transmission power (ICTP) adjustment, where nodes with more information use higher transmission powers than those that are less informative to share their target state information with the neighboring nodes.
In this module, a wireless network is created. All the nodes are randomly deployed in the network area. Our network is a mobile network, nodes are assigned with mobility (movement).Source and destination nodes are defined. Data transferred from source node to destination node. Since we are working in mobile network, nodes mobility is set ie., node move from one position to another.
When assessing the security of systems such as nodes, one major problem comes from the absence of widely accepted adversarial models giving a precise description of the attacker’s goals and his capabilities one such model for secure machine learning and discussed various general attack categories. Our work does not fit well within because our main goal is not to attack the learning algorithm itself, but to recover one piece of secret information that, subsequently, may be essential to succesfully launch an evasion attack.
Revisited Closely related to the points discussed above is the need to establish clearly defined and motivated adversarial models for secure machine learning algorithms. The assumptions made about the attacker’s capabilities are critical to properly analyze the security of any scheme, but some of them may well be unrealistic for many applications. One debatable issue is whether the attacker can really get feedback from the system for instances he chooses. This bears some analogies with Chosen-Plaintext Attacks (CPA) in cryptography. This assumption has been made by many works in secure ma-chine learning, including ours.
For performance evaluation we use the following graph Packet delivery ratio Throughput Delay
Recent work has accurately pointed out that security problems differ from other application domains of machine learning in, at least, one fundamental feature: the presence of an adversary who can strategically play against the algorithm to accomplish his goals. A few detection schemes proposed over the last few years have attempted to incorporate defenses against evasion attacks.
Our attacks are extremely efficient, showing that it is reasonably easy for an attacker to recover the key in any of the two settings discussed. We believe that such a lack of security reveals that schemes like kids were simply not designed to prevent key-recovery attacks. however, in this paper we have argued that resistance against such attacks is essential to any classifier that attempts to impede evasion by relying on a secret piece of information.
We have provided discussion on this and other open questions in the hope of stimulating further research in this area. The attacks here presented could be prevented by introducing a number of ad hoc counter measures the system, such as limiting the maximum length of words and payloads, or including such quantities as classification features. We suspect, however, that these variants may still be vulnerable to other attacks. Thus, our recommendation for future designs is to base decisions on robust principles rather than particular fixes.
In this project, we present ECD, an Efficient Code Dissemination protocol for wireless sensor networks. Compared to prior works, ECD has three salient features. First, it supports dynamically configurable packet sizes. By increasing the packet size for high PHY rate radios, it significantly improves the transmission efficiency. Second, it employs an accurate sender selection algorithm to mitigate transmission collisions and transmissions over poor links. Third, it employs a simple impact-based backoff timer design to shorten the time spent in coordinating multiple eligible senders so that the largest impact sender is most likely to transmit. We implement ECD based on TinyOS and evaluate its performance extensively. Results show that ECD outperforms state-of-the-art protocols, Deluge and MNP, in terms of completion time and data traffic.