2.4.3. DPoS

DPoS is a consensus process that allows shareholders to vote on the nomination of witnesses [52]. In DPoS, the main objective is to minimise energy waste and accelerate transaction times. As a result of the overall block generation process, this consensus mechanism operates much more quickly than PoW consensus. In DPoS, each stakeholder is allowed to cast one vote per share; they can cast additional votes when they own additional coins. Moreover, the witnesses are rewarded for producing blocks and penalised for failing to do so, such that they are not paid and are voted out of office. To complete the instructed task, witnesses should receive the largest number of votes from random stakeholders. A stakeholder also votes on the restructuring of the delegates and adjusts the network, which will be reviewed by the stakeholder before a final decision. Although DPoS was designed to increase transaction efficiency and overcome the constraints imposed by many other consensus mechanisms, it has significant shortcomings. The network is not sufficiently decentralised due to a large number of validators. A centralised system may serve as a focal point for random intruders due to its centralised nature. DPoS is susceptible to 51% attacks because an attacker will convince stakeholders to give them 51% voting power in a 51% attack [40]. Asymmetric agreements are also vulnerable to other types of attacks, such as long-distance, DDoS, P + epsilon, Sybil and balanced attacks.

This work investigated the fact that the three main consensus systems, which are susceptible to several attacks, have significant weaknesses. As a result of their vulnerability, digital transactions are at a high risk of being attacked. Table 1 summarises the results of our analysis. The 51% attack can exploit all three consensus mechanisms, making it desirable for attackers, particularly for PoW where achieving the required hashing power is cost-effective.


**Table 1.** Vulnerabilities of consensus mechanisms.

The following is a brief description of several severe attacks. However, the focus of this study is on 51% attacks. Long-range attacks are the result of a weak model of subjectivity [53]. This form of attack is similar to the 51% attack. It appears to fork the chain from the genesis block [54] rather than confirming the sixth block. This type of attack occurs very rarely in Bitcoin, but it can be damaging when demonstrating stakeholder consensus (PoS) and delegate stakeholder consensus (DPoS). Assuming a PoS consensus scenario in which the invaders begin with a limited number of coins shortly after the genesis block, their chain versions can be privately mined to carry out the attack. Given that they have a small stake, they will generate a limited number of blocks at the beginning and then generate a longer chain. PoS does not specify a threshold for chain lengthening, so the chains can become extremely long. The P + Epsilon attack is a method of exploiting the dominant strategies of the participants in the network. PoW-based blockchains are usually vulnerable to this type of attack [55]. When attackers give participants a pay-ou<sup>t</sup> in order to gain an advantage, a payoff matrix is used where the dominant tactic facilitates the achievement of the attacker's objectives. As a result of the attack, the participants do not receive any compensation, and the attacker receives the entire amount. This key statistical finding is based on an ad hoc selection model.

### *2.5. Hybrid Approaches Related to 51% Attacks*

The term 51% attacks refers to situations in which an attacker has 51% of the hashing power. As part of this attack, a private blockchain is created and completely disconnected from the actual chain edition. It is later introduced to the network as a real chain, which allows for a double-spending attack [47] Additionally, given that blockchain policy follows the most extended chain rule [56], if attackers gain 51% or more of the threat, they will push the longest chain by convincing network nodes to obey their chain. However, 51% of computational power is not strictly sufficient, so double spending is still possible if an attacker has less than half of the computational power [48]. The odds of success are low. A blockchain attack becomes increasingly expensive when the entire network acquires

increased hash power. A cryptocurrency with a high network hash rate may also be resilient to 51% attacks. To overcome 51% attacks, several studies and developments are being conducted. Researchers have proposed mixing proof mechanisms to eliminate 51% attacks. For PoW to be applied to a working network, the attacker must gain more than 50% of the processing capacity and more than 50% of the network wealth. This task is highly challenging for a user. In addition, the total cost should be considerably lower than the profit that an attacker might earn. The attacker's costs are much higher than the benefit in this form of a hybrid network.

Additionally, this hybridisation implements other security measures to counter the attack [28]. Different studies and innovations have recommended different prevention methods, but they have several limitations in common. Komodo [57] introduced dPOW consensus, which takes a snapshot of the blockchain every 10 min and stores it in the blockchain. One of the more recent developments implemented by Horizen at Zen Coin is to delay the block in order to slow down the creation of blocks [29]. Casper and Decred provided a second hybrid PoS consensus using a BFT model with a two-thirds vote mechanism in the first 50 networks. This voting process is independent, resulting in unintended delays and an inconsistent block interval [30]. An alternative hybrid consensus algorithm was proposed by the authors in [31], namely, fork-free hybrid consensus with versatile proof-of-activity and the hybrid PoW-PoS-PoA algorithm. The authors introduced a technique where all PoW chains are created simultaneously and submitted to a committee for review. The committee determines and approves the most robust chain as the main chain. A weighted calculation amongs<sup>t</sup> the committee members determines which chain is the best. On the basis of PoW power and PoS capacity, the weight of each committee member is calculated. This algorithm can also mitigate 51% attacks. However, other issues may affect the blockchain. One of the primary issues is determining how to distribute newly produced block rewards [27]. Another issue is that the interval between block generation is often inconclusive, which is directly related to the recently created currency [49]. Table 2 summarisesvarioushybridapproachesandothersolutionsdiscussedintheliterature.


 **Table 2.** Comparison of the proposed hybrid mechanism and other hybrid solutions.

> In conclusion, this study proposes a hybrid algorithm that combines PoS and PoW mechanisms to provide a fair mining reward to both the miner and validator. By maintaining a precise block generation interval with difficulty adjustment in power mining and a likelihood measurement according to the stake's mature stake balance, the system not only resolves the 51% attack but also provides stakeholders and investors with a uniform distribution of mining rewards.
