Disaster Management System Aided by Named Data Network of Things: Architecture, Design, and Analysis.

Disasters are the uncertain calamities which within no time can change the situation quite drastically. They not only affect the system's infrastructure but can also put an adverse effect on human life. A large chunk of the IP-based Internet of Things (IoT) schemes tackle disasters such as fire, earthquake, and flood. Moreover, recently proposed Named Data Networking (NDN) architecture exhibited promising results for IoT as compare to IP-based approaches. Therefore to tackle disaster management system (DMS), it is needed to explore it through NDN architecture and this is the main motivation behind this work. In this research, a NDN based IoT-DMS (fire disaster) architecture is proposed, named as NDN-DISCA. In NDN-DISCA, NDN producer pushes emergency content towards nearby consumers. To provide push support, Beacon Alert Message (BAM) is created using fixed sequence number. NDN-DISCA is simulated in ndnSIM considering the disaster scenario of IoT-based smart campus (SC). From results, it is found that NDN-DISCA exhibits minimal delay and improved throughput when compared to the legacy NDN and existing PUSH schemes.

Authentication of Smartphone Users Based on Activity Recognition and Mobile Sensing.

Smartphones are context-aware devices that provide a compelling platform for ubiquitous computing and assist users in accomplishing many of their routine tasks anytime and anywhere, such as sending and receiving emails. The nature of tasks conducted with these devices has evolved with the exponential increase in the sensing and computing capabilities of a smartphone. Due to the ease of use and convenience, many users tend to store their private data, such as personal identifiers and bank account details, on their smartphone. However, this sensitive data can be vulnerable if the device gets stolen or lost. A traditional approach for protecting this type of data on mobile devices is to authenticate users with mechanisms such as PINs, passwords, and fingerprint recognition. However, these techniques are vulnerable to user compliance and a plethora of attacks, such as smudge attacks. The work in this paper addresses these challenges by proposing a novel authentication framework, which is based on recognizing the behavioral traits of smartphone users using the embedded sensors of smartphone, such as Accelerometer, Gyroscope and Magnetometer. The proposed framework also provides a platform for carrying out multi-class smart user authentication, which provides different levels of access to a wide range of smartphone users. This work has been validated with a series of experiments, which demonstrate the effectiveness of the proposed framework.

Major-Minor Long Short-Term Memory for Word-Level Language Model.

Language model (LM) plays an important role in natural language processing (NLP) systems, such as machine translation, speech recognition, learning token embeddings, natural language generation, and text classification. Recently, the multilayer long short-term memory (LSTM) models have been demonstrated to achieve promising performance on word-level language modeling. For each LSTM layer, larger hidden size usually means more diverse semantic features, which enables the LM to perform better. However, we have observed that when a certain LSTM layer reaches a sufficiently large scale, the promotion of overall effect will slow down, as its hidden size increases. In this article, we analyze that an important factor leading to this phenomenon is the high correlation between the newly extended hidden states and the original hidden states, which hinders diverse feature expression of the LSTM. As a result, when the scale is large enough, simply lengthening the LSTM hidden states will cost tremendous extra parameters but has little effect. We propose a simple yet effective improvement on each LSTM layer consisting of a large-scale Major LSTM and a small-scale Minor LSTM to break the high correlation between the two parts of hidden states, which we call Major-Minor LSTMs (MMLSTMs). In experiments, we demonstrate the LM with MMLSTMs surpasses the existing state-of-the-art model on Penn Treebank (PTB) and WikiText-2 (WT2) data sets and outperforms the baseline by 3.3 points in perplexity on WikiText-103 data set without increasing model parameter counts.