In this paper, we propose and evaluate a new in-memory index structure for efficient processing of range queries over moving objects. Traditional spatial index approaches suffer from the need for frequent updates and therefore result in poor performance in moving object environments. To reduce the number of updates, many existing approaches use a linear function to represent the movement of objects, but for numerous real applications, the movements of objects are too complicated to be represented as a simple linear function. The index structure we propose has two levels. The lower level consists of a number of cluster nodes, each of which is in charge of a group of moving objects and can dynamically adjust its location and covering range according to the movement of the objects. The upper level makes use of a grid index to support querying evaluation. With our approach, no assumption is required about the maximum velocity of objects and it is not necessary for objects to move according to some well-behaved patterns.