Article Preview
TopIntroduction
M2M is a communication paradigm that provides ubiquitous connectivity between devices along with the ability to communicate autonomously without human intervention.
Market size forecasts show a great potential for the M2M market, which is expected to grow rapidly in the next few years. It is caused by a number of factors, including the widespread use of wireless technologies, lower prices for M2M modules and economic incentives.
M2M acts as a stimulating technology for the practical implementation of the Internet of Things (IoT). In this context, IoT is considered as “a global network of connected devices that have identities and virtual personalities working in intelligent spaces and using intelligent interfaces to communicate in social, environmental and user contexts”. This IoT vision represents a future in which billions of everyday objects and environments will be connected and managed through a series of communication networks and cloud servers.
The number of peripherals is growing exponentially around the world, providing quality services to the user. The Internet of Things makes it easy to share information between users, devices and applications that are located in different locations around the world. This is a good start-to-start new research on the possibility of using portable devices to solve real optimization and machine learning problems.
It is expected that an increasing number of messages will generate a large volume of data, thus increasing the number of attacks for malicious users due to the openness, distributed nature and lack of control over the entire IoT environment.
The purpose of this article is the development and software implementation of the steganographic method based on a genetic algorithm (GA) that allows to modify the existing algorithms of steganographic protection in such a way as to increase their noise immunity without introducing additional redundancy, which would reveal the inclusion of additional information and reduce the security of M2M interacting systems.
Novelty of the article: a steganographic model is implemented to hide data, including large amounts of data, in graphic images with the help of the created modification of a genetic prototype algorithm.
The classic error-correcting coding algorithm is based on the addition redundancy. However, adding extra redundancy to the message being transmitted may not be an acceptable price. Therefore, there is a problem in reducing the effects of possible distortion of characters in the transmission without making additional redundancy in the message. The proposed approach does not introduce additional redundancy in the transmitted message and, accordingly, does not lead to the correction of errors. Nevertheless, it allows you to decode the received message (possibly distorted) into the closest one as applied to the whole ensemble of messages. In this case, the gain reaches 7σ or more. The initial prototype algorithm is described in (Kudryashova & Adzhemov, 2018a). Building an algorithm for estimating the effective coding of a source when converting signals in various metric spaces, the study of its properties and optimization are devoted to the works (Sineva & Batalov, 2013; Fenchuk & Sineva, 2014; Fenchuk, Batalov, & Sineva, 2014). Genetic coding algorithms were developed to improve noise immunity without introducing additional redundancy (Batalov & Sineva, 2015a, 2015b; Fenchuk, Sineva, & Bott, 2016; Yakovlev & Sineva, 2014; Fenchuk & Sineva, 2015).
Having carried out an analysis of GA, it can be noted that the above algorithms do not take into account the distribution of errors in different symbols of binary code. At the same time, the encoding devices and the channel can lead to the fact that the probability of distortion in different bits is different, or the consequences of such distortion are different, from the point of view of message processing. For example, in the method of the least significant bit (LSB) the distortions in the last bit lead to the distortion of the built-in hidden information and are the most critical. Accounting for this circumstance at the stage of coding will reduce the impact of the bitwise distortion from the point of view of minimizing the distance between the sent symbols and decoded in the metric space of the source.