 In the context of network management, we know that simple network management protocol is used to oversee the performance of a network. And if there is a requirement for changing the behavior, then some variants of it can be incorporated that also implement certain policies as laid out by the system administration. There is a term known as self-organizing network. It is a proper noun because it was coined by Duis Telekom that is an operator in Germany in the next generation mobile network alliance. We are going to look at various aspects of it and we will see what is unique about SONs and how these can be understood in order to do some research into it because it is forming the basis of self-learning networks which alter or change their behavior as the network progresses with a varying load of users. We know that networks became intelligent over time from being configurable to getting automated and some programming was introduced in software-defined networking for network function virtualization and this led to an automatic execution of certain policies that the system administration or the network operators would like to. This has taken us to a completely different level known as autonomic computing or autonomic networking where the system becomes independent. This independence is enough for a network to do fault performance accounting and security management at its own. The goal however is to take it to a level where the network becomes self-aware but we know that consciousness is not a virtue of computing systems so that dream or that goal largely remains theoretical. There are some self-organizing features which are available to networks for quite some time especially in the context of telecom networks, telcos, which have a typical architecture of a base station, base transceiver, mobile switching center, network operation center, administration and management center so in there the mobile station and the base transceiver have a power control feature which is pretty much an automatic if not automated system. Then the handover soft or hard takes place through a certain rationale or an algorithm which is prescribed by the network operator. Similarly a customer is recognized by subscriber identification module known as SIM which contains the wallet and the payment details of a certain user and the service package to which the customer has subscribed. So we are aware that this kind of self-organization or self-management was already there in telecom networks however as the networks grew the applications grew the number of users became very large their interaction with the networks became very complex so some drivers or some motivations led to the evolution of self-organizing networks. The most obvious one is since manned networks are difficult to have no error because the likelihood of human error is always there. So in order to avoid the laborious man induced errors some kind of automation would be required that immediately would result into cost reduction and the overall complexity that the humans cannot readily fathom or comprehend holistically can be delivered to the machine. So it means self-organizing networks are value-added network management orchestration mechanisms. This would need even more automation even more programming so we are going to look at some intelligence for the networks to become more enabled to organize themselves without human involvement. This also means that some kind of over the air programming or TAP or remote method invocation mechanisms should also be incorporated. This would require some communication connectivity and computation and their blend so as I said earlier Jewish telecom in the next generation mobile network alliance came up with this son initiative back in 2007 which was adopted. Let's look at a brief history of it. The operators which were known once as the walled gardens were making a lot of money but as the competition grew as their profit margins reduced they came up with an initiative to identify the use cases where human element of error should be reduced. So these were expected to be automated because there was a lot of human effort involved these were error-prone and the nature of such use cases was complex so these were put up to vendors because the hardware actually comes from the original equipment manufacturers and certain third generation partnership project standardization bodies to look into standardizing self-organizing solutions. This is the basic of background or historical perspective to sons today we see sons already deployed because in 2023 most of the telecom networks already have adopted sons our purpose is to look at the continuous need for self-organizing that eventually leads to self-awareness which is highly cherished or here in goal. Today we see plug-and-play deployment of base stations in telecom operators for that the process is simple the onsite installation is an expectation on the end of the service providers to the vendor has somehow enabled the device to configure itself and do some software management activity in an automated manner when it is turned on. An example is the automatic neighbor relationship configuration where a base transceiver in 2G networks or node B in 3G or E node B in 4G becomes aware of the neighbors which it has to possibly reach out for load balancing for interference control and noise reduction etc. So this results into overall better management and the overall network quality is experienced by a user also increases so son was put up as a formal standardization activity in which essentially there are two parts the first one is it was recognized that the way a typical telecom network provider operates has to be automated because it is only through automation that self-organization can be implemented so the workflow execution system is the first and integral part of self-organizing networks this includes a policy mechanism that is policy storage policy definition policy enforcement points where the policy actually can be configured on the fly on the network elements. The self-organization capability is implemented as a closed management loop compare this to an open management activity where human involvement is something as the last call but for network elements for devices that is the network management and the overall umbrella management this closed loop management loop implies that the devices would not be involving any human intervention are going to do some kind of device to device configuration orchestration communication to come up with an adaptive behavior that is the need of the network at that particular moment let's look at the manual network operation we know that everything starts with planning so once we have the plan this plan is put into operation and the continuous monitoring of the operation is done through performance fault and configuration management because on ground we have hardware like racks where we have telecom infrastructure like node B E node B media gateway with their own set of alarms alerts and error reporting mechanisms kind of sensing and then actuation means scripts which run in response to those predefined set of errors and alarms and failures. Then this is reported to to the network optimization mechanism that keeps learning this and shares this feedback with the human being so this net manual network operation involves human being compare this to closed loop automation where we have a network planning that of course starts with the human being but then it becomes pretty much an automatic closed loop where we have the network plan with its own set of policies however since we have now identified two essential elements the workflow automation and the self organizing network capability through policy enforcement here we can see that the network plan is implemented in true letter in spirit through workflows and the policies the policies are enforced on ground into the into the network elements the domain administrative domain and the overall network as as an autonomous entity so the human effort which was done once at the ebb and issue time or at the starting time is now shifted to to more abstract level of planning but on ground the system is running almost at its own so the experiences which are now being learned in network operations are now being shared into back into the workflows and policies and these workflows and policies are going to be becoming part of the network plan and the network plan is going to be implemented after storing it in some kind of repository so you can see that the network plan along with the policy on how to manage and run the operations on the network is a closed loop activity this lecture and the subsequent lectures to follow have been adopted from a very beautiful book by John Wiley back and it was published back in 2020 2012 and the first author he's he's an accomplished network engineer to my last understanding when I knew of him he was here at Nokia but he started off his initiative at Doesh Telecom so we are going to refer to this book in all the associated modules related to self-organizing network