Characterizing patent thickets by patent triples via the multiplex network of standard essential patents

Patent thickets (PTs) significantly increase transaction costs and may hinder subsequent technological development. However, three existing approaches do not adequately capture the defining characteristics of PTs, and most are unsuitable for large-scale analysis or for providing a macro-level view of PT structures. This study proposes a network-based framework for identifying PTs in information and communication technology (ICT). Standard-essential patents (SEPs) are selected because they share some similar characteristics with PTs. Using 105,502 extended patent families from all ETSI-declared SEPs, we construct the multiplex network PCN-PSCN-PJCN-PPCN and calculate it further to obtain the network PIN. Then, we define t ψ -structure and apply the Louvain algorithm to obtain clusters which are considered as PTs. The framework strictly meets the five defined elements of PTs. The results demonstrate the effectiveness and rationale of the framework, with 11 structurally cohesive and technologically interpretable PTs found. The PTs exhibit substantial internal connections and closed triples, with No 1 PT being the largest and No 2 PT showing the strongest local overlap. Their themes cover resource management, beamforming, device-to-device communication, system information handling, and adaptive transmission. The 5G phase contains the largest number of PTs, whose themes center on improving network coverage, reducing communication latency, enhancing bandwidth utilization, and supporting application scenarios including machine-type communications, the Internet of Things, smart homes, and connected vehicles.