Advancements in the internet as a technology over the years have come with a great challenge of malicious network attacks. The detection of such attacks which are known as network intrusions has become a very well-researched area. This study is aimed at contributing to the research carried out in the detection of network anomalies. The study carried out a comprehensive literature review on the subject of network anomalies and carried out experiments on the WSN-DS network anomaly dataset using artificial neural networks, with CNN-LTSM as the choice of artificial neural network algorithm. Experiments were carried out on five different attack type labels in the dataset, and the results of the experiments showed promising performance accuracy with results of accuracy anomaly detection performance accuracy of 98% and precision and recall rate of up to 99% accuracy. The results of the experiment put CNN-LTSM as a suitable artificial neural network algorithm for network anomaly detection in real-time scenarios.