 One of the first requirements for self-organizing networks is self-configuration. Configuration lies at the heart of optimization for faults, that is reducing faults, improving the performance, having better account and audit of the network and to keep the system secure. Self-configuration is something that is at the essence of self-organization. We'll see how the requirement for self-configuration was understood and we'll look at an interesting scenario of commissioning and radio-configuration. At times when certain telecom operator is providing services to certain user base with increasing market size with changing customer requirements, there's a need to either roll out new network elements or provide changes to the network elements. This would entail replanning and reconfiguration because planning and deployment go hand in hand. The problem becomes very complicated because the number of network elements at the base stations is a very important entity which is central to the network design. So, if some change is required at the network element, it would incur a lot of cost if it is not automated. For LTE, that is long-term evolution, starting from 3G literally to 4G, 5G and 6G now, the number of cells have been increasing continuously and the size of the cell has been reducing. The commercially off-the-shelf equipment is now being incorporated from the provider side to bring the cost low. So, it means that it is becoming very hard to provide manual configuration. Add to this the central OANM activities that is optimization, administration and management of the operations, administration and management. If it is done centrally, it means the overall control for security was consistent. But now that the network elements are being installed in diverse numbers with diverse range of variety in hardware and software, so security becomes a concern. So, self-configuration now has to incorporate security as a default feature as well. Let's look at a scenario where base station has to be installed in a certain network proximity. Base station in pure 2G jargon talks to the base station controller, but in 3G, 4G and beyond, the network elements have been rebranded, but essentially the concept remains the same. If a certain base station has to be deployed, it means it's a new network element. It has to be installed and restarted with minimum human intervention. For that, the self-configuration dictates an interesting observation. That is, in the preparation or planning phase or in the rollout or deployment phase, these two phases were connected or were placed in tandem. But now the operators and the manufacturers joined their hands to eliminate the interaction between the two and merged these two into one self-configuration activity. Now, the network element once bootstrapped has minimal off-the-shelf configuration, which is available in it on some EE prom, which contains the default operator settings. Once the network element is placed on site, it's part of an understanding between the manufacturer and the operator to allow the network element to map to the existing operator settings. This leads to independent commissioning, which involves a mix of planning and deployment. In this figure, it has been taken by Professor Seppo Hamilton from an IEEE paper again. It was published in 2007. In here, the classical phases, which were more of silos, if you may call them, included preparation at the factory level, then on-site installation using field engineers, then on-site commissioning by the software engineer. Now, these steps have been obviated by using software. So, auto-configuration or self-configuration essentially implies that the moment a BTS or a Gateway base station is installed in a certain network, using some kind of dynamic host configuration protocol or its variant like auto-configuration, the device will have some policy that it will implement because the device is aware of the workflow execution and its exact state when it has to trigger the interaction with some network device. So, we see that we achieve automatic connectivity upon commissioning. So, where is the human element then? The human element has been obviated and you can see in this diagram in the box below that the auto-configuration allows to gain the ease in terms of time, in terms of the human resource required and in terms of errors that humans would make in such complex things. So, the gain or flexibility gain was achieved and we see that the overall rollout time essentially improves. From the operator perspective, the device is ready to use on-the-fly. From the vendor perspective, it doesn't have to send its own field engineers and software engineers to meet the requirements of the operator. Let's look at another interesting scenario where we are interested in radio-configuration. Radio-configuration implies adjustment of radio parameters like transmit power, SI, NR, signal to interference, noise ratio adjustment, the thresholding of key parameters. So, once we are done with connectivity after commissioning, it complements this radio-configuration. It complements the configuration parameters used for commissioning which is part of the firmware which was downloaded in the commissioning phase. So, what happens is the radio-configuration parameters are prescribed as at some default value or some kind of negotiation even can take place between the base station or the base transceiver with the base station controller or home in order to be talking to the media gate where the higher entity in 3G and beyond networks. The network elements adapt the configuration parameters to current state of network deployment making it more suitable to adjust to the changing channel behavior. We see that the design, build, operate and maintain states that we know are very much part of any life cycle of a network. We are already through with auto-connectivity and security setup. Auto-commissioning was also done but in dynamic radio-configuration phase, the device adjusts to the network conditions and the channel model and it goes through a continuous monitoring and reporting mechanism to optimize its behavior and if there is an error to heal itself. This means that these use cases dictate some kind of automated behavior where human involvement is not there. So this is what we call plug-and-play as well. The reference is Sepohemilenin, a long-term evolution self-organizing network.