Since the emergence of sensor oriented applications, sensor data management has become a very active research domain. On the one hand, sensor networks provide distributed processing of queries on the sensors, on the other hand, data stream management systems are systems which deal with data stream comming from sensors. Recently, some hybrid approaches have also appeared which aim to integrate these two different kinds of solutions. This paper presents these three groups of proposals and gives their approach to deal with issues such as scalability of sensor systems, heterogeneity of sensors and processing of complex queries on data stream comming from sensors.