IEEE Consumer Electronics Magazine - January/February 2022 - 43

In order to provide better services, consumer
electronics products usually have a network interface
to achieve interaction, remote operation and
remote backup, etc., and service providers can use
this interface to maintain and update software,
firmware, and even functions. From the consumer's
perspective, consumers do not need to know
the signaling process behind the service; they only
need to focus on enjoying the required service.
Hence, many consumers often neglect to improve
information security so that their important information
will be exposed to risks through the Internet,
allowing malicious user to easily retrieve these
transaction data through some attack tools. At this
moment, if the data are not protected by encryption
mechanism, the stolen transaction data will be
easily deciphered immediately.3
A commonly used encryption way in the consumer
electronics is symmetric encryption,
because this kind of encryption has the characteristics
of extremely fast operation. Adopting of symmetric
encryption allows all consumer electronics
products to complete interactive services more
smoothly. However, this kind of encryption way
has a fatal flaw that encryption rule is too linear.
This is because that a more regular key generation
mechanismresults in a higher probability of generating
weak keys.6 It meansthataslongasthe malicious
user has a strong computing machine, the
ciphertext of the consumer will be deciphered easily
through brute-force attack.4;5
In order to overcome the defect of symmetric
encryption, enhancing randomness is a necessary
way. In this article, a novel metaheuristicbased
key generation framework is given that
considers 1) increasing randomness of generated
keys; 2) decreasing the number of occurrences of
weak keys; and 3) reducing the calculation time
as much as possible to find the best solution.
Based on the abovementioned major goals, we
collate the contributions of this study as follows.
1) An optimization problem is defined for key
generation of symmetric encryption.
2) Designing a metaheuristic-based framework
for generating the keys as well as clearly
describing the missions of these architectural
elements.
3) A key similarity model is defined as the fitness
function, which is specifically for metaheuristics
to evaluate the quality of the keys.
January/February 2022
4) Three different types metaheuristics are presented
for key generation, which are singlesolution-based
algorithm, population-based
algorithm, and swarm-intelligent algorithm.
5) This research provides a faster and relatively
secure encryption method for the consumer
electronics industry, especially in the field of
IoTs such as smart home, smart city, and so
on, which has more powerful advantages.
BRIEF DESCRIPTION OF ENCRYPTION
Cryptography is a profound knowledge and has
a history of decades. Encryption methods can be
roughly divided into three types: symmetric
encryption, asymmetric encryption, and hashbased
encryption. The common symmetric
encryptions are data encryption algorithm (DES),
triple data encryption algorithm (3DES), and
advanced encryption standard (AES). DES is one of
the oldest encryption technologies. In response to
the performance of early computers, the design of
DES does not require a large amount of computing
resources, and the computing speed is very fast.
Although the security is not high, it is very sufficient
for the early service requirement. 3DES is an
extension of DES for enhancing the security, but it
must consume a lot of computing resources, and
has a more computing time due to a longer key
length. AES is amore advancedmethod, which has
the longer key length so that AES can obtain more
security than DES even 3DES. Moreover, AES only
takes a little bit more time than DES, but it is secure
than 3DES.
The abovementioned encryption methods
belong symmetric encryption so that it is more
suitable for single user scenarios since its encryption
and decryption use the same key. If a multiple
users service platform wants to use encryption, it
is more appropriate to adopt asymmetric encryption
due to the use of different keys. RSA is a classical
asymmetric encryption, which can be used for
general encryption or digital signature so it is a
widely used method today also. Because the difficulty
of a generated key from RSA depends on the
complexity of digital decomposition, it can provide
certain security as long as the length of the key is
enough. However, for this reason, it must pay
more computing resources for digital decomposition.
Digital Signature Algorithm (DSA) is also a
common asymmetric encryption. The encryption
time of DSA is very short but can provide good
43
IEEE Consumer Electronics Magazine - January/February 2022

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - January/February 2022
