CalNode is a prototype Cognitive Network Access Point (CogNet AP), which has the unique capability to observe and learn from the network traffic in order to optimize itself. Unlike traditional devices which require significant amount of network planning prior to deployment, a cognitive network system based on CalNodes can be deployed with no prior channel planning. Even with no planning, the cognitive ability of these devices enable them to converge to the optimal network configuration over a period of time. These devices collect, compact, repositorize, and analyze wireless network traffic in order to extract crucial spatio-temporal network traffic patterns. The information gained from the spatio-temporal network traffic patterns will be used to reconfigure the network elements for optimal performance. In addition, information gathered by CalNodes will be shared among themselves for improving system efficiency. CalNodes can be used in either centralized  or autonomous mode. In the autonomous node, decisions are taken  within a device with the help of local information. However, in the centralized mode, a collection of CalNodes are controlled by a centralized CogNet controller in order to achieve network-wide optimal configuration. In the emerging heterogeneity of wireless networking environment, cognitive networking capability of CalNodes help design new network forms that can achieve higher network capacity while minimizing the effort needed to deploy them.