Software reliability is undoubtedly the most important aspect of software quality. One can define the same as the chances of computer program operating without any failures in a particular environment for a specified period of time.
 
In order to improve upon software reliability one has to consider all the different stages of a program's development, namely requirements classifying, designing, programming, testing and runtime evaluation. One should be aware that the study of theoretical reliability of software is concerned to a large extent with the concept of correctness, a mathematical field in computer science which is primarily based on automata theory.

One cannot expect a program to work as desired if the developers/designers of the same are not made aware of the program's desired behavior in advance. Developers should have the ability to determine in sufficient detail the software's desired behavior in correspondence with its development. However, the idea of perfect detail might seem attractive, but it is equally impractical in practice, since the desired behavior tends to change frequently. This is because the possible range of desired behavior is determined from the actual attempts or even the failed attempts to do the same. 

A high-level design has the following objectives:

• It distinguishes the problems of architecture, overall program concept and structure from the actual problems of coding, thereby solving the problems of data processing.
• Applies extra constraints to process of development by narrowing the scope of the smaller software components and removing variables that increase the chances of programming errors.
• Provides a program template that includes the specification of interfaces, that can be simultaneously shared by distinguished teams of developers who are working on disparate parts enabling them to know in advance how their contributions will interface with those of the other teams .
• It specifies the program independently of the implementation languages, thereby removing language-specific biases and limitations

Programming: The evolution of programming languages can be viewed as a way of getting the computer to carry out more and more of the work. Without a deep and thorough understanding of the program's overall structure and functionality, it would be impossible to detect errors in the program. Therefore, the use of better and enhanced languages should considerably reduce the number of errors Using an improved language impart the software an even greater amount of abstraction. Inventions, such as statement, sub-routine, file, class, template, library, component, help in arrangement of a program's part. Moreover, it also enables an exact control over the shape and use of data elements. 
 
Testing: A high quality of software testing confirms that the product stands up to its requirements. Testing includes the following:

• Unit Testing
• Functional Testing
• Performance Testing
• Failover Testing
• Usability Testing

It is important to use testing early in the development cycle in order to ensure a high quality in the products. For instance, the practice of test-driven development, under which tests are written before the codes to be tested, is common in the extreme programming to ensure quality.

Runtime: Determining runtime reliability is quite similar to tests. But it goes beyond  simply confirming the behavior of the software to the evaluation of qualities.