MEASURE THE LEVEL OF SENSORY PROCESSING PROBLEM IN CHILDREN WITH AND WITH OUT AUTISM IN CLASSROOM ENVIRONMENT

Background of the Study: Sensory processing is the process in which all sensory stimuli integrate and enable person to respond according to the situation. There are number of sensory issues that either parents or teacher do not identify on early stages and various times it also cause behavioral issues or academic issues with in typical children too. The study aims to assess the sensory vulnerability of children with and without autism in the school environment and to identify the sensory issues. Methodology: This was a cross-sectional study conducted in Karachi, Pakistan. Total 86 children recruited and divided into two groups included typical and atypical children. Using a standardized questionnaire-SPM (sensory processing measure) distributed in schools and completed by teachers. Result: Autism affects both mainstream children and children with autism, with sensory process subscales showing significant differences p value (< .005) for t- test.  Autism children have higher sensory process vulnerability, particularly in social participation and praxis variables, compared to mainstream children’s. Conclusion: The study reveals that school initiates sensory stimuli, affecting children's vulnerability, even in special schools, despite providing structured environments for atypical children.

Computational Intelligence Based Secure Clustering Techniques for Vehicular Adhoc Networks

VANETs, an application of MANETs, enable ITS using IEEE 802.11p standard which is in favor of DSRC specifically designed for WAVE scenario. VANETs establish communication among vehicles (V2V) and road side infrastructure (V2I); while V2I communication using IEEE 802.11a/b/g standard. In VANETs vehicles, road side entities disseminate FSAMs about road conditions and other vital circumstances to ensure safety and avoid losses of precious lives and property. As in VANETs system vehicles move with high speed, so due to high mobility environment and topology also changes with time. In VANETs system accurate and on time delivery/reception of FSAMs is highly important to withstand against maliciously inserted security threats affectively. Hence, there is no optimum routing protocols which ensure on time delivery of FSAMs to destination. Due to frequent alteration in VANETs topology path failure, inter vehicle distance change and malicious node penetration may also result. So absolutely optimum protocols for secured delivery of packets exchange is still challenging. Clustering for VANETs is extremely beneficial but stability of existing clustering algorithms for VANETs exhibit poor robustness due to their dynamic nature. In this thesis, a new clustering algorithm is presented for VANETs by the name of moth flame optimization-driven, reproducing the social behavior and hunting approach of moth flames in designing proficient clusters. Due to the random range of VANETs, stability is a major area of research which has gained much attention. The main idea of presented algorithm is extracted from the living routine of moth flames. Presented algorithm permits well-organized communication by generating the amplified number of clusters and their unsupervised working make it as intelligent. Intelligent Clustering using Moth Flame Optimization (ICMFO) scheme is accomplished for determining and optimizing the clustering issues in VANETs; the primary focus of which is to enhance the stability in such networks. ICMFO is then validated by comparison with two other existing variants of Particle Swarm Optimization (PSO), i-e; Multiple-Objective Particle Swarm Optimization (MOPSO) and Comprehensive Learning PSO (CLPSO) and one existing scheme of Ant Colony Optimization (ACO) known as Ant Colony Optimization Based clustering algorithm for VANET (CACONET). Simulation results demonstrate that ICMFO is providing optimal results in comparison to existing techniques. It is also cleared from the proposed work results of different researchers, that there is no such protocols that is best suited for clustering as well as security implication in VANETs. Different routing schemes have different conduct performance metrics. In our thesis we concentrated and inspected different routing protocols. We have also presented a new security based scheme named ARV2V; and compared its results with existing techniques which are Trust and Logistic Trust in terms of TCE, EED, ALD and NRO. The scheme has presented security implication in our clustering based scheme ICMFO. In terms of TCE, ARV2V is 11.6% and 7.3% efficient than LT and Trust respectively. In terms of EED, we found ARV2V 57.6% performance 5.2% better than LT, also Trust schemes met 52.4% more delay than LT.Similarly, in term of ALD ARV2V provides 29.7% and 7.8% more stable link duration than Trust and LT respectively, however LT has 21.9% proficient ALD than Trust. ARV2V protocol have 27.5% and 14% lesser load than Trust and LT respectively in terms of NRO, while Trust has approximately 13% more NRO than LT.