The Internet of Things (IoT) is utilized in many sectors for different purpose. However, sharing the huge data to the far distance is a difficult task because of signal loss and malicious events. So, the Delay Tolerant Network (DTN) is introduced in the communication field to transfer the data to far distance. But, the DTN is vulnerable to attack because of the selfish nodes and data traffic. To overcome this problem the current research has designed a novel African Buffalo DTN with Risk Packet Jump (ABDTN-RPJ) mechanism to improve the communication channel by predicting the selfish and misbehavior node in an earlier stage. Finally, the designed paradigm is implemented in the Network simulator and the comparison results have been proved the efficiency of the proposed model by reducing the delay and drop rate.