Unconstrained Online Optimization: Dynamic Regret Analysis of Strongly Convex and Smooth Problems