PhD Literature Survey Writing services

CriticalReview:

In recent times the web based development is gradually increased thus it makes a traffic to access a certain same resources thus it makes to down the servers or unable the access the server. It occurs mainly by generating a malware to reduce an access through fake logins, etc. And one of the main issues to address is to ensure data availability along with taking into account the well-known confidential, integrity and availability (CIA) triad. Here time delay and power consumption is also referred for loss of performance efficiency this also be a major factor on basic of data transmission, thus there is need to be fix all issues/errors in a random manner to ensure the data secure and a proper data transmission.

MethodicalReview:

There are three types of intrusion detection methods typically distinguished, signature based, anomalybased, and specification based. Mostly the combination of methods or more attains a high gaining as a hybrid approach. Thus signature based IDS proposed to detect if the attack signature is already stored in database along this, anomaly based detection recognize to malicious behavior thus this combination provides an excellent detection and routing attacks. By this methods the attacks are tacked the behavior regularly for modules. It helps to deploy certain lightweight modules in networks. Thus this approaches does not requires any modification/sensors is required. And the non–network side channel attack is another method to exploit the hardware of IoT devices.

Procedural Review:

To provide a secure resource it need to follow certain common procedures, likewise to mitigate the UDP Flooding attack in IoT network that the count based rate limiting mechanism. A node in Contiki able to limit the rate at which it sends an ICMP error message helps to reduce the consumption of bandwidth thus the error message after overall N packets are dropped. Here the counter counts message and dropped packets then exemplified the logic with certain algorithm. To Simulate and evaluate the mitigation scheme on basis of transmission power and CPU processing power of the victim node. To provide the communication ability to resource by implementing Contiki TCP/IP through UDP flooding self-programmed on a number of client modes thus by this process consumption of power and efficiency based on its allocated time.

Descriptive & Hypothetical Review:

DDoS attacks have been, and more powerful versions of them are still being used as potential methods to disrupt the networks’ security. In several visualization-based detection that the models have been developed for network security of traditional wired networks, this idea has not much in securing Wireless sensor Networks (WSNs), which predominate today’s network systems across the globe. This is because of these networks are dynamic, can be built with minimal resources, and are cost-effective. While researchers are trying to tap into the full potential of WSNs, the attackers have also taken advantage of these sensor networks’ dynamic nature. They have launched many types of DDoS attacks. Thus, in this thesis work, a visualization-based DDoS attack detection model for WSNs has been proposed.

Conclusion:

In this approach, a dataset containing non-attack and UDP attack data was generated to identify optimal features for attack detection like energy consumption, power consumption, packet loss, delay, CPU time, and idle time. In normal conditions, the sensor nodes use low amount of reserves for task completion. However, in under attack situations, they consume higher than normal amounts of these resources. Not all conditions under which a sensor node uses more energy than usual can be suspected as attack conditions. Thus another feature called Received Packets (number of packets received by a node per unit time) which helps distinguish between the attacker and non-attacker nodesthe clustered data points are then visualized onto a parallel coordinate graph to identify the unique patterns they form and discern attack points from non-attack points. The malicious points lie on one edge of the cluster axis, whereas the benign points lie on the other end. The comparative performance analysis of both the clustering algorithm determines that mini-batch k-means has an improvement of 10%over the k-means algorithm. This reveals that the developed system is accurate and straightforward in detecting UDP flood attacks in WSNs and has made a huge contribution to the field of network security