Fast and reliable wireless communication depends mainly on the available frequency bandwidth．Cognitive radio has the ability to reliably and autonomously identify unused frequency bands.With cognitive radio technology promising to utilize spectrum bands．The question still remains；can it utilize the spectrum bands efficiently with the increasing scarcity of the spectrum?　　The growing number of the wireless users brings challenges such as degradation of the quality of service of the real-time applications caused by the primary user activity andinterference from increasing number of Cognitive radio users．　　When primary users appear in the spectrum band，CR users need to move to a new availableband，resulting in a temporary communication break．To solve this problem，that multiple noncontiguous spectrum bands are assumed．they are simultaneously used for thetransmission in the CR network．This method can create a signal that is capable of high data throughput and also immune to the PU activity．Even if a primary user appears in one of thecurrent spectrum bands，the rest of them will maintain current transmissions．　　Real-time applications need to have more reliable and time invariant communication channels to satisfy strict service requirements，such as delay constraints and sustainable rates．　　However,how to maximize the total network capacity is still a crucial problem．The proposed practical spectrum decision scheme aims at improving the quality of service by reducing delay time and data loss which occurs due to a temporary transmission breakdown．Whenever real-time application CR users fail，to get an available channel with their sustainable rate to continue transmission．They yield a high data loss which makes the QOS more unreliable．The proposed assignation queuing criteria allows real-time users to queueslots for less than one as they wait for available channel to continue transmission．When compared to the tradition spectrum decision，they are an improved data loss．　　The Qos were investigated with the primary user activity performance in MATLAB simulations．When implementing the queuing order criteria we assumed that there was no interference from primary user activity．Generally,the queuing order criteria showed reduceddata loss．The proposed assignation queuing order criteria having the least data loss ascompared to the random and reversed.