APIs are the main mechanism for integrating heterogeneous components in modern software systems.API documentation typically treats each endpoint as autonomous, leaving out information about other endpoints that produce or consume input or output data, which is useful in rapidly orchestrating valid call sequences.This makes the development, testing and maintenance of systems that use APIs to internally implement business logic, or integrate with third-party API-based services, more complex.This paper addresses this gap by introducing static and dynamic analysis methods to automatically discover and document inter-endpoint dependencies; static analysis uses existing API documentation, while dynamic analysis is based on logging API usage.Both methods are integrated into RADAR, a tool for extracting and visualizing graphs, to illustrate such inter-endpoint dependencies through exchanged data.Case study on the PayPal API demonstrates the value of adding dependency graphs to API documentation to support rapid development and documentation of software systems.

As cyber-attacks become more and more sophisticated, sharing information that helps organizations design and implement efficient defense measures, is of critical importance.Such information can be shared using any service available, such as plain-old mailing lists, forums, etc.More mature systems use standards that facilitate the structural and semantic organization of information about cyber threats, which enables both automated processing and interpretation of such info, such as indexing, cross-referencing, updating, and more.However, even systems sharing cyber-attack info are themselves vulnerable, not only to typical and easily detectable attacks such as DoS, but also to content poisoning.Implementing such systems using decentralized architectures such as Blockchain, could overcome many deficits of centralized cyber-threat-info sharing systems.This paper presents the specification, design and implementation of such a decentralized system using two popular standards for cyber threat intelligence sharing, namely STIX for representing and TAXII for sharing such info using a REST API.The system, implemented on Hyperledger Fabric, faces the challenge of adhering to standards designed for a centralized world, and offering a transparent way for implementing all the backend, on a Blockchain.

Wireless sensor networks are becoming quite popular and gain ground to a wide range of applications. A new integrated system for forest fire management has been implemented and put into pilot operation. The system offers services ranging from in-situ forest temperature measurements, to optimization, risk assessment, and decision support for wildfire management. The elements of the system provide real-time functions to detect, monitor and predict the evolution of forest fires along with other operational logistics services. A state-of-art self-organizing expandable fire sensor network is used for real-time in-situ measurements. The network uses sensors deployed camouflaged on trees, which implement a self-organizing RF data network that broadcasts the forest temperature distribution using redundant local control nodes and dynamic-re-routing so as to maintain contact with the central server. This paper presents the requirements, considerations and experience gained in the development and pilot implementation of the in-situ self- organizing network for forest temperature and wild fire monitoring. The pilot implementation took place at "Nea Penteli" forest, located approx 20 km NE of Athens.