We consider an optimal asset allocation problem for a defined contribution pension fund in the accumulation phase before the participant retired. It is assumed that the prices of $n$ risky assets suffer a common jump phenomenon because of the impact of market systematic risk. So the prices of the risky assets are modeled with dependent jump-diffusion processes. In addition, we suppose that the income process of the participant satisfies a jump-diffusion process before she retired. To minimize the expected square loss between the pension fund account and the expected investment target at the time of retirement, the stochastic dynamic programming method is adopted, and the closed-forms of the optimal investment strategy and the value function are obtained, a verification theorem and its proof are also provided. Finally, through some numerical examples, we find that the jump parameters in the price process of risky asset and participants' income process, the earnings preference of pension fund managers and the expected investment objective process all have significant influence on the optimal investment strategy and the value function.