Information security is based on cryptography, steganography or combination of both. Cryptography is used to collapse and lock the meaning conveyed by the secret message by making it understandable only to the users with a specific key to unlock the secret. The steganography and watermarking intentionally make an unintended message visible to all that serves as cover to suppress the guess on the presence of secret behind the medium.
Lightweight encryption schemes designed for crypto applications on resource-constrained embedded devices are known as Light Weight Cryptography (LWC). They are often used for data security in IoT devices.
There is a conference paper on lightweight cryptography " A recent review on lightweight cryptography in IoT" , available on
https://ieeexplore.ieee.org/abstract/document/8058307/
The aim of steganography is to hide information imperceptibly into a cover, so that the presence of hidden data cannot be diagnosed,.
I think there is almost no use for steganography in the IoT. There is a paper :
file:///C:/Users/hp/Desktop/Nada%20thesis/Security%20and%20Privacy/2015%20Internet%20of%20Things_%20Securing%20Data%20using%20Image%20Steganography.pdf
There are two challenges in IoT Steganography (in my opinion):
1. Lightweight identification (or autentification, but not so clear). Stegano techniques can be more useful because light requirement (versus crypto), we can use literally bites for identification.
2. Modifications control. IoT devices and firmware can be compromented or modificated and we have attacks with false firmware (IoT botnets, APT etc.). Watermarking of network stegano are both useful for these tasks.
I suppose, that the main challenges are the following:
1. To make IoT device inaccessible to intruders. As Vyacheslav Zolotarev stated IoT devices could be compromised and to do this for steganography is much easier, than for cryptography.
2. Find suitable containers. In processors world the data containers are much better optimized, than ones used at application level, so you should find one big enough.
3. Provide robustness. Sent data (especially technical one - used in transfer protocols, for instance) could be modified or excluded from message, when it finally reaches its destination.
4. Programming some steganography algorithms for certain RISC processors (such as multimedia processing algorithms) could be a hell.
Also I share Peter Breuer's interest about 32-bit RISC processors. And if specify the processors family considered by you it would help us to provide better answers.
Apart from already mentioned challenges I find following further challenge:
Lack of a generic cryptographic algorithm standard considering wide variety of devices with variety in hardware,architectures.
- Badges
- Science topic
- Similar topics
- Instrumentation Engineering
- Instruments