Autonomous Network Management System (ANMS) is a network system that is self-governing. The network requires minimal or no human involvement on its operation. It is capable of managing and regulating itself. It enables automated second-by-second decision-making by distributed components that exchange information fast and accurately. Autonomic network re-emphasizes efforts of the networking group on more significant tasks of outlining general application performance goals and controlling the accomplishment of objectives against application-based Service Level Agreements (SLA’s). It has the capability to self-learn, self-configure and self-optimise towards delivering the objective set by its administrator. Such a system can manage thousands of computer users, sites and applications. It has traffic management patterns that constantly change with the changes in initiatives. This makes it easy to manage applications (such as social networking, voice, video and peer-to-peer collaboration), initiatives (such as cloud computing, unified communications and data centre consolidation), and real time user activity. An automated system comprises of numerous different sub networks with joined routing logic and common routing policies. Every single subnetwork has an assigned international unique identification number comprised of 16 digits, known as the AS number. These identity numbers help an automated system to exchange exterior routing information between two neighbouring networks.
Autonomous network systems communicate routing information within using interior gateway protocol but use border gateway protocol while sharing information with other autonomous systems. It is capable of delivering a fully automated application visibility, wireless area network, dynamic wireless area network among others over an enterprise’s network regardless of the enterprise’s size and technology. Because of the complexity of networks as they grow large and carry different data formats, there arises a need for service levels (SLA) to be close to or at 100 per cent. ANMS is like the human nervous system, which controls all the aspects of breathing, heartbeat and the rest without the conscious participation of the brain. It is integration of autonomic computing conceptions and network management ideologies that create ANMS. The complexity in the network brought about by integration of networks of different types and protocols in the internet. The heterogeneous landscape of networks has created the need to have the highly skilled network engineers’ work on highly complex technologies thereby creating the ANMS.
According to IBM (2005), the autonomous network management goals have been categorised as self-optimising, self-protecting, self-configuring, and self-healing. In self-configuring, the system automatically and dynamically adds fresh nodes as they get substituted on the network. This leads to the scalability of the network. In self-healing, the system discovers, diagnoses and takes actions to prevent any interference, as well as rectification of any occurring faults. Such action may lead to the system and its subsystem changing its own state or influence changes in other elements of its environment. Faults that are not easily detected in manual systems (such as presence of virus in the system) are automatically corrected. In self-optimisation, the system monitors and controls its resources towards ensuring optimal functioning as per the pre-defined requirements. Self-optimisation assists in creating high standard of service for the system users and external clients.
The network deals with large amounts of data collected over an extended period. The system can use its self-optimisation feature to assess and anticipate the need for additional capacity. This feature can also anticipate, recognise and guard against dangers from anywhere in the network system. By self-protecting, the system automatically controls access to the network resources; it only allows the right to users with the privilege to access the right information thus stopping unauthorised users or hackers from attacking the network system. This feature is always facilitated by its analysing technique; it analyses the data collected from recurrent unapproved logins and any suspicious communication patterns on the network. For instance, if the system detects an attack, it may shut down, restart or lock itself automatically depending on the level of threat. This self-protecting feature reliably enforces safety, thereby minimising the managerial security expenditures.
The main goal of an autonomous network management system is to improve the performance and dependability, at the same time reducing the costs of network management. To achieve this objective, the system uses a variety of automated techniques. This ranges from agent-based methods relying on explicit models and ontology to developing techniques dependent on gossip protocol, swarming algorithm or other biologically inspired work. The ultimate goal of ANMS is to overcome the ever-increasing needs and complexity in networks, especially the internet, and to enable more development from the current state, ensuring that there is no limit to a network’s usability. The use of the internet has been an extraordinary success as it grew exponentially over the years. The number of users is still increasing, with this trend expected to continue, proper management methods will be core in delivering quality and reliable services. It is evident that the human capability to operate and manage these networks is being exceeded. This means that automated network management systems are essential.
Autonomous network management system can be used to manage enterprise networks with up to tens of thousands of users. With emergence of cloud computing and other increasing IT capabilities, presentation of resources as services over the internet for on demand access at any time anywhere. Networks are responsible for the delivery platform for services. It is thus essential for the network operator to guarantee reliability of the delivery platform so that both consumers and the service provider can depend on the network. Several aspects of cloud computing such as extremely large numbers of services, virtualised infrastructure, thousands of users, and the lack of direct control services create significant challenges on the networking management for service delivery.
The ANMS offers a solution to such problems and ensures proper and reliable service delivery. An autonomic network management system ensures that the operating region it encompasses likely failure scenarios and does not exacerbate the occurring problem. For instance, when a link between autonomous systems fails the internet’s border gateway protocol replies by eliminating routes that depend on that (broken) link. It then selects a new route as required and announces the resultant changes to its neighbours. This process will take a relatively short time. The ANMS can be used to control such organisations as internet service providers, government bodies and institutions that have many diverse networks operated under the umbrella of a single entity for ease of management. It can also be used in large enterprises that have the requirement of many smaller networks, which are geographically separate but operated using a related environment.
The autonomic systems are self and context aware. These features refer to perception and cognitive reaction towards an event or a condition, relevant to the same intelligent node or respective to the environment. This context awareness is foundation for the rest of its operational features. These autonomic networks are fundamentally composed of a set of interacting autonomic entities. Each autonomic entity is responsible for a set of managed resources, consume services and offer services to human or peer autonomic entity. These autonomic entities manage their own internal behaviours and their interaction based on pre-established or on-line agreements. These entities can re-negotiate the terms of agreement in order to describe a new practice of interaction or collaboration.
A network management system is a setup of utilities designed to enable a network administrator to monitor and safeguard health and smooth running of the network. It monitors both the hardware and software constituents of the network. These utilities typically record data from remote points on the network and relay the data to a central administrator. The main difference between autonomous network management system and normal network management system is that it contains the proficiency to monitor large amounts of data, applications and users among several things that normal networks cannot do. The ANMS utilises the concept of autonomic computing, which was an inventive and kicked off in the year 2001 by IBM, with an eventual goal of producing a self-managing network that would overcome the network complexities.
On the strength of the automated system, an ANM system possesses the ability to manage itself, including self-configure, self-protection, self-optimisation, self-healing and self-troubleshooting. It re-establishes connectivity without human involvement to achieve the network objectives. It has higher reliability compared to the normal network management systems. They have improved response times when request for change on the networks is faster than respondent to which reduce the network downtime. As compared to the normal networks in which appeal to the network alteration has to be collected, check on the problem record and then respond back. The ANMS also reduces the skill requirement of a network engineer. This is because its network tasks get encoded within the system automatically as opposed to the normal system where an engineer performs the management task. It thus reduces the amount of skills necessary to accomplish these network tasks, whereas a normal network management system requires an engineer for run be available all the time.
The third strength of the ANMS is the automation of operations that are manual in the ordinary network management systems. ANMS automatically respond to the requests instantaneously based on the information collected from the system. This decreases the manual labour and time required to correct critical situations.
In the normal system, human network management uses the ticket system to track down problems, and it is time consuming. ANMS enables better management control of a network. The rapid response helps to solve problems in the network. The quality of service includes error rates, delay, bandwidth and availability. Normal network management systems require time to respond to requests as all the occurrences gather information and then check with problem record before correcting the problem at hand. This tends to reduce the quality of service as well as the network availability, if the problem has to stop working. The next strength is reduction of cost. ANMS reduces maintenance costs as all errors get solved within the system automatically without or with minimal human intervention.
In the normal network management systems, skilled engineers are required to sort out errors and faults manually. ANM System is fault-tolerant compared to the normal network management systems. ANMS rectifies its faults by using the proactive and active solution. With the proactive solution, network faults are detected before the network is degraded and this solution keeps the network available and running at all time. Normal networks use reactivity most of time to overcome faults in the network. This may mean network shutdown in order to solve faults. This makes the network unsteady and not available for some time. Reactive solution communicates on manual prompting compared to the proactive, which gets prompted automatically.
Autonomous network management system has weaknesses, as discussed below. The first one is high upgrading costs. The ANM Systems are self-governed and base their work on the set objectives and goals. When changes occur in the manner of operating the business, or there occur changes in the business policy, the system rejects the new policies and deals with them as threats. This requires a system upgrade for new changes to take effect. This may mean an upgrade of the entire system or re-programming which will involve significant costs. Another weakness is interaction of components. ANMS uses exact devices, language and platform originating from a common vendor. This indicates that any component to be used in the system has to come from that vendor. Any device to be used in the system has to be compatible with the original vendor’s technology. In some cases this means replacing the entire system. In other cases the components may be unavailable or too expensive. The last weakness of the ANMS is network disruption. The ANMS adapts dynamically to the changes in its environment. This process might cause disruption in the user’s devices already in use when implemented during busy or working hours.
This paper will discuss five critical evaluation methods used in autonomous network management system against normal network management systems. These include configuration management, fault management, performance management, accounts management and security management. Starting with configuration management, ANM Systems are always self-managing in their configuration to any device added to their network. They use mobile agents which are software agents and are able to move between locations and make their own independent decisions, according to the changes happening in their environment. The ANMS use mobile agents to self-configure devices added to their network without need for a network engineer. This process does not require the prompt or presence of an engineer. On the other hand, the normal network management system uses SNMP (simple network management protocol) architectural model. This SNMP uses agents that reside within the managed network device to collect data that are then passed on SNMP to the network engineer, for an engineer to manage the network devices remotely. The network engineer then manages the devices manually. It can be said in conclusion that automated systems are efficient on the basis of configuration, when compared with the normal network management systems. This is because the automated system configures itself and allows the system to continue without affecting its end users in cases of change in the system environment.
As for the fault management, the ANMS work with artificial intelligence technology to sort out any fault capable of causing network degradation proactively, without any human intervention on the system. The fault could be hardware or software failure that is capable of causing the entire network failure. The autonomous management system has the capacity and ability to diagnose such a fault and use the appropriate procedure to hint the cause of the fault and automatically rectify that fault. The management system uses network agents to alert the engineer about failure of hardware or software. The engineer then gathers information about the network problem. This can be extremely problematic at times, particularly if the fault is complex or the network is large. After the engineer has collected enough information, he/she then diagnoses the problem manually. This procedure can be slow and time consuming, causing down time of the network for a long time.
When it comes to the performance management, the normal network management system deploys performance management techniques; the monitoring tool detects and scrutinises the network trend for overloading, for instance. The system then sends an alert message to the engineer asking him/her to find a solution before the system network threshold exceeds the value set. The Automated System is, however, able to manage and monitor itself on the performance of network. It dynamically tunes itself to respond to the prevailing workload and improve on the overall performance.
About security management. The automated system protects the entire network dynamically by the use of its self-protecting technique. The system blocks any unauthorised access to file and network resources. Its security system monitors users who do not have the authority to access the network resource. They are thus denied access if they try accessing unauthorised files in the network. This is normally done by collection of previous unauthorised logins and environmental changes which are used to determine potential threats. When it comes to the normal network systems, it has a security monitoring tool that limits the user’s ability to access unauthorised resources such as sensitive system files. When there is an attempt to access unauthorised resource or file, the operating engineer gets notified by the security management software tool.
As for the account management, in the normal network management system, the engineer is responsible for allocating network resources. He/she allocates resources and has the responsibility to understand utilisation of network resources so as to allocate them appropriately. In the Automated Network Management System, the system dynamically allocates essential network resources to the end users. Because the resources are automated to suit the end user, the network becomes more productive and easy to utilise.
In conclusion, from the evaluation that was discussed in this paper, the ordinary network management systems can be used in relatively smaller networks than the automated systems. This is because the normal network management systems better suit to the smaller networks, such as medium to small business enterprises. The Automated Network System, on the other hand, is suitable to a large network system that is still expanding and has large network demands. The ANMS is suitable to large business enterprises and institutions of higher learning (colleges and universities). Normal network management systems have become error-prone, especially when the network is large and complex. The network managers find it difficult, costly and time consuming to manage such a network, especially when the network is massive.
The future environment will certainly be more complex in terms of networking requirements and management systems. This is evident as humankind explores the heights of technological advancements. This comes with a need to embrace technology in every venture, from businesses to education and communication systems. As humanity struggles with the ever-increasing complexity of the management process, the future will have more autonomic systems that will be entirely independent, without any need for human intervention. It is for this reason that people should embrace technological advances as they make work easy to manage. They also ensure high levels of efficiencies that cannot be achieved by human brain.