SummaryNew mobile devices offer multiple network interfaces to allow the users to connect to the best available network. The heterogeneous networks can provide better internet connectivity to the users by means of vertical handover. The handover must be triggered at a suitable point of time to avoid mobility issues such as unnecessary handovers and handover ping‐pongs. The network selection during handover is usually done using classical multi‐attribute decision making (MADM) methods. However, ranking abnormality is one of the prominent issues of the classical MADM methods. To address these challenges, a graph theory and matrix approach (GTMA) with Euclidean distance is proposed for vertical handover in wireless networks. GTMA is used for the selection of the appropriate network and Euclidean distance is utilized for the handover triggering. The simulation results reveal that the proposed method has eliminated the ranking abnormality issue. This proposed technique without triggering has also reduced the number of handovers up to 75.61%, 85.71%, and 66.67% as compared to the traditional MADM methods such as AHP, GRA, and TOPSIS respectively. The use of Euclidean distance for handover triggering has further reduced the number of handovers of the proposed technique as well as traditional techniques for all the traffic types.