Bitcoin's decentralized architecture protects users’ privacy by leveraging pseudonyms at the data layer and mitigating unlikability at the network layer. The evicting-filling attack [11] takes advantage of flaws in network processes, such as the common connection pool, which breaks this promise. This research overcomes the evicting-filling assault and finds various flaws such as its limited scalability, vulnerability to changes in the network in real time domain, and lack of full countermeasures. We recommend using the Adaptive Isolation and Randomization Defense (AIRD) framework to reduce the likelihood of addresses being used. It combines connection pool isolation, randomized eviction timing, and real-time network monitoring. Experimental results on Bitcoin Mainnet nodes demonstrate AIRD's efficacy, with a 98% reduction in successful address linking attempts compared to the baseline. Our findings broaden the attack's potential to new digital currencies like Cardano and Ethereum 2.0, and we suggest avenues for further investigation into creating an environment that protects users’ privacy fully.