Defending against DoS (denial of service) attacks has become a great challenge, especially for institutions that provide access to their services in the public network. State-of-the-art identity concealing tools and vast number of computers connected to the network require ensuring appropriate means for entities at risk to enable defence from the particular type of threats. This article presents a concept of user authentication in IP communication. The concept consists in providing the receiver with the possibility to determine sender՚s identity at the Internet layer level. This provides both the capability of defence against DoS attacks and possibility of utilizing the presented model over existing Internet network, which is directly responsible for transmission. The authors hope that the concept is a significant step in the perception of public network data transmission.

This article describes security mechanisms used by 3rd-7th layers in OSI/ISO network model. Many of commonly used protocols by these layers were designed with assumption that there are nointruders. Such assumption was true many years ago. The network situation has been changed for last few years and we should realize that some properties of existing protocols may be abused. Moreover, we should exchange some of them or create new versions. There are some methods and guidelines concerning secure programming, but there is also lack of guidelines about creating secure protocols. Authors see the necessity of such guideline and this article is an attempt at analysing existing solutions and selecting some universal and important patterns.