The importance of managing events has increased steadily over the last years and has reached a great magnitude in science and industry. The reasons for that are twofold. On the one hand, sensor devices are cheap and provide event information which is of great interest for a large variety of applications. In fact, sensors are ubiquitous in modern life. Nowadays, RFID tags are attached to goods and parcels to allow an easy tracking. Numerous weather stations are providing up-to-date information about temperature, pressure, and humidity to allow for precise weather forecasts worldwide. Lately, mobile phones are equipped with various sensor devices like Global Positioning System sensors or acceleration sensors to increase the applicability of the phone. On the other hand, reacting on events has become an increasingly important factor especially for business applications. The occurrence of a system failure, a sudden drop in the stock exchange, or a missing parcel can cause huge costs for the company if their appearance is not handled properly. As a consequence, detecting and reacting on events quickly is of great value and has lead to a change in the design of modern software systems, where event-driven architectures and service-oriented architectures have become more and more important. With the emerging establishment of event-driven solutions, complex event processing (CEP) has become increasingly important in the context of a wide range of business applications such as supply chain management, manufacturing, or ensuring safety and security. CEP allows applications to asynchronously react to the changing conditions of possibly many business contexts by describing relevant business situations as correlations over many events. Each event corresponds either to a change of a business context or the occurrence of a relevant business situation. This thesis addresses the need to cope with heterogeneity in distributed event correlation systems in order to i) reuse expressive correlation and efficient technology optimized for processing speed, ii) increase scalability by distributing correlation tasks over various correlation engines, iii) allow migration of correlation tasks between heterogeneous engines and security domains, and iv) provide security guarantees among domains in order to increase interoperability, availability and privacy of correlation results. In particular, a framework called DHEP is presented that copes with such requirements.