REAL-TIME V2V COMMUNICATION FOR INTELLIGENT
TRANSPORTATION SYSTEMS
Introduction
Rise in vehicles necessitates intelligent transportation system (ITS)
VANET enables vehicle-to-vehicle (V2V) communication to realize
ITS
Provides safety, traffic monitoring, route computation etc.
Literature Review
Early protocols like AODV used flooding for path discovery
Some require storing and carrying packets till next hop found
Grouping vehicles into clusters with cluster heads for inter-group
communication
Most use predefined static data instead of real-time traffic data
Research Gaps
Improving reliability and minimizing delay and retransmissions
Developing robust city protocols resilient against interference
Handling temporary network disconnectivity
Reducing congestion due to excessive route discovery packets
Minimizing unnecessary packet overhead
Proposed Methodology
Traffic-aware optimized greedy geographic routing
Active probing and passive monitoring for traffic and reliability info
Piggyback street reliability data to avoid separate broadcasts
Recalculate routes at intersections using piggybacked data
Greedy forwarding to anchor points deemed reliable based on data
Key Highlights
Efficient V2V communication is key for ITS applications
Routing must adapt to highly dynamic vehicular environments
Traffic-aware optimization can improve reliability and reduce
overheads
Next-gen protocols needed to handle real-time data and changing
conditions
DIAGNOSING DEPRESSION USING DIFFERENT APPROACHES
Introduction
Depression a major health concern, affects mental state and
communication
Questionnaires and scales used for clinical diagnosis
Machine learning techniques explored recently for automatic
diagnosis
Challenges
No specific symptoms, difficult to detect in early stages
No standard laboratory tests, subjective clinical assessments
Multifactor disorder, needs information from multiple sources
EEG-Based Diagnosis
Non-invasive, cost effective, correlates brain activity with depression
Techniques use EEG features like entropy, fractal dimensions etc.
Classifiers: SVM, Random Forest, Deep Learning models
Accuracies up to 99% reported on EEG datasets
Multimodal Diagnosis
Combines multiple modalities like EEG, MRI, speech, video, texts
Provides more information than single modality approaches
Fusion of features and classifier outputs
Improves accuracy and robustness compared to single modality
Conclusion
Automatic diagnosis can aid clinicians and enable early detection
EEG provides good discrimination capability
Multimodal fusion recommended for handling complexity of
depression
Key Highlights
Machine learning has potential for automatic depression screening
EEG a promising modality, provides brain activity patterns
Multimodal fusion improves performance compared to single
modality
Future work needed in multi-modal fusion and interpretable models
